Full Disclosure

$29.99 (G)

View Excerpt as PDF (302KB)

ONE

Governance by Transparency

THE NEW POWER OF INFORMATION

On September 12, 2000, Masatoshi Ono, the chief executive of leading U.S. tire manufacturer Bridgestone/Firestone, faced a panel of senators and a battery of television cameras in a packed hearing room. The senate panel was investigating mounting deaths from a mysterious series of auto accidents in which tires blew out without warning, causing vehicles – many of them Ford Explorer SUVs – to roll over. Addressing the senators and the room full of victims’ families, auto safety advocates, and industry representatives, as well as a nationwide television audience, Ono uttered words that no CEO wants to say: “I come before you to express my deep regret and sympathy to you, the American people and especially to the families who have lost loved ones in these terrible rollover accidents.”¹

   The Firestone scandal remained national news during the summer and fall of 2000 because auto companies and tire makers had failed to inform the public about deadly risks. Documents from Firestone/Bridgestone and Ford indicated that both companies had been aware of a pattern of fatal accidents caused by a combination of tire tread separation and top-heavy SUVs but had done nothing to alert drivers. Bridgestone/Firestone executives knew that its plant in Decatur, Illinois, where most of the problem tires were made, had long had quality-control problems. When the count was finally complete, 271 people had been killed in accidents involving problematic SUV design and defective tires.²

   The public, however, learned about these problems only by chance – and only after many of the deaths and injuries. In early February of 2000, Houston station KHOU-TV reported that lawsuits claimed that exploding Firestone tires and associated Explorer rollovers had caused thirty deaths. It took another six months for Bridgestone/Firestone executives to acknowledge the problem and recall 6.5 million tires, the largest tire recall since the 1970s.³

   More troubling, the Senate investigation revealed that the problem was larger than a limited number of defective tires. In the 1990s, many people bought SUVs because they thought they were safer than smaller cars. The Firestone/Explorer revelations showed that, to the contrary, SUVs were more likely to roll over than other cars – and some SUV models were much more prone to roll over than others. That was important because rollovers remained the most deadly auto accidents, accounting for nearly a third of auto fatalities in the United States even though they represented less than 4 percent of all accidents.⁴ Nonetheless, information about which SUVs were prone to roll over – like the facts about the unusual Firestone tire blowouts – remained locked in company files.

   As reports of deaths and injuries mounted, congressional committees debated what action to take. Regulators had authority only to mandate recalls and impose modest fines on automakers and tire companies for safety defects. States could prosecute officials for criminal negligence, of course, and injured passengers could sue for damages. But such actions would not reduce the likelihood of future deadly accidents.

   Circling around contentious issues concerning how to mandate safer design, Congress instead legislated targeted transparency. The Transportation Recall Enhancement, Accountability, and Documentation (TREAD) Act, approved in November 2000, required auto companies for the first time to give car buyers the facts about each model’s rollover risks so that they could make their own safety choices.⁵

   The idea was not just that the public deserved better information. It was that the power of information would create a chain reaction of new incentives. Armed with new rollover ratings, buyers would choose safer cars. Confronted with declining sales of the most top-heavy SUVs, auto companies would improve design. Safer design would save lives and prevent injuries. The new law thus made transparency into a precise policy tool.

   Information had new power because policymakers did not stop at simply placing facts about risks in the public domain – where they could easily be lost in the cacophony of new-car hype. Instead, policymakers required that information be presented in a format that was designed to be user-centered. They distilled the complex probabilities of rollovers into simple five-star ratings based on government tests of each new model (see Figure 1).

   In a few seconds, car buyers, regardless of their math or language skills, could compare risks and identify rollover-prone models. A five-star vehicle, with a 10 percent or less chance of rolling over in a single-vehicle crash, was much safer than a one-star vehicle, with a 40 percent or more chance of rolling over. Buyers could also customize information to suit their needs. Anyone interested in more detail could delve deeper into narratives and Web site links. In 2005, Congress made the policy even more user-centered. A new law required that information be presented by September 2007 where car buyers most needed it, on showroom new-car stickers.⁶

   The policy also included an interesting built-in mechanism intended to increase the chances that transparency would be sustainable. It required that ratings become more accurate over time. The initial scores would be based on simple mathematical modeling of rollover propensity combining each model’s center of gravity and track width. But the law required that safety regulators also work toward a road test that would more accurately mimic real-world driving conditions, and it directed the National Academy of Sciences to study possible tests and required regulators to consider the academy’s recommendations. As a result, regulators instituted a new test in 2004 that combined modeling with driving maneuvers.⁷

   Congress added other disclosure-based incentives. The TREAD Act re- quired tire pressure monitoring sensors by 2003;⁸ safety regulators required automakers to disclose information on customer complaints and other early indications of safety defects;⁹ and new labels made it easier for car owners to see if their tires had been recalled.¹⁰

   This, then, was the concept: government would use the power of information to drive better choices by car buyers, which in turn would improve vehicle designs and reduce risks. But would it work?

   Five years after the release of the first set of rollover ratings, the answer appeared to be yes. Initially, SUV models had widely varying rollover rates – and most performed poorly. In 2001, thirty models received only one or two stars, meaning that they had a greater than 30 percent chance of rolling over, while only one model (the Pontiac Aztek 4-DR) earned a four-star rating, meaning that it had a less than 20 percent chance of rolling over. By 2005, however, only one model (the Ford Explorer Sport Trac) received as few as two stars, while twenty-four models earned four stars.¹¹

   Transparency also created pressures that ended a generation of industry lobbying against a rollover safety standard. The national attention that rollover accidents received in 2000 and the new star ratings spurred auto- makers to accelerate their introduction of stability-control technology. By 2005, 20 percent of new vehicles were equipped with sensors that triggered corrective braking, compared with fewer than 5 percent in 2000. Voluntary adoption of new technology changed the political dynamic. In 2005, Congress approved the Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (SAFETEA-LU), which directed regulators to issue minimum performance standards for auto rollovers.¹²

TRANSPARENCY INFORMS CHOICE

In recent years public attention has focused mainly on struggles over broad transparency in government – President Clinton’s championing of a broader public right to know and President George W. Bush’s controversial moves to increase government secrecy, for example. Few have recognized that a second generation of targeted transparency has been rapidly gaining ground.

   Instead of aiming to generally improve public deliberation and officials’ accountability, targeted transparency aims to reduce specific risks or performance problems through selective disclosure by corporations and other organizations. The ingeniousness of targeted transparency lies in its mobilization of individual choice, market forces, and participatory democracy through relatively light-handed government action.

   Since the mid-1980s, scores of targeted transparency policies have percolated up through the political system in the United States – usually without any awareness by their creators that they were participating in a more general innovation in governance. After a deadly chemical accident in Bhopal, India, killed thousands of people, Congress required in 1986 that manufacturers tell the public about the toxic pollutants they released – factory by factory and chemical by chemical. After scientists confirmed that unhealthy eating habits were contributing to millions of deaths from heart disease and cancer each year, Congress required in 1990 that food companies inform the public about the levels of fat, sugar, and other nutrients in each can of soup and box of cereal. After a series of revelations about the surprising frequency of serious medical mistakes, Congress considered proposals in 2000 to require hospitals to inform the public about such mistakes, and several states required hospitals and doctors to tell the public their mortality rates for specific procedures. After the corporate accounting scandals of 2001 and 2002, Congress required that public companies improve their financial disclosure.

   Targeted transparency policies have also been crafted to improve the fairness and quality of public services. In response to continuing concern about financial institutions’ discrimination against inner-city borrowers, Congress strengthened requirements in 1989 and 1992 that banks report on their mortgage loans according to the race, gender, and income level of borrowers in each geographical area they serve. In response to continuing concern about the quality of public schools, Congress required in 2001 that school performance reporting demonstrate school improvement as a condition of federal aid.

   Each of these laws wrested from the files of corporations, nonprofit organizations, or public agencies some of the facts that executives would often like to keep confidential – information about the risks they create and about flaws in the quality of goods and services they provide. Each offered sunlight in a format that poor performers would most like to avoid – in labels, reports, or Web sites that allowed consumers, investors, employees, and community residents to compare products and practices.

   Though the problems they address vary widely, the idea behind all these new laws is the same. A generation of research by economists and political scientists has shown that markets and deliberative processes do not automatically produce all the information people need to make informed choices among goods and services. When hidden risks or service flaws create serious problems for the public at large, the government can help reduce those risks or improve services by stepping in to require the disclosure of missing information.

   Why is government action needed? Three reasons: First, only government can compel the disclosure of information from private and public entities. Second, only government can legislate permanence in transparency. Third, only government can create transparency backed by the legitimacy of democratic processes.

   The core characteristics of targeted transparency policies are also the same. It is hard to imagine that nutritional labeling, school performance ratings, and corporate financial reporting have much in common. Yet all targeted transparency policies include these characteristics:

•   mandated public disclosure

•   by corporations or other private or public organizations

•   of standardized, comparable, and disaggregated information

•   regarding specific products or practices

•   to further a defined public purpose.

•   Information users perceive and understand newly disclosed information

•   and therefore choose safer, healthier, or better-quality goods and services.

•   Information disclosers perceive and understand users’ changed choices

•   and therefore improve practices or products

•   that in turn reduce risks or improve services.

   While new in its broad information-age applications, targeted transparency is not a new idea in governance. In 1913 Louis D. Brandeis, the “people’s attorney” and later Supreme Court justice, wrote in Harper’s Weekly that “sunlight is…the best of disinfectants.” Brandeis recommended new laws to require public companies to disclose their profits and losses in order to stop insider deals that deceived investors. He pointed to an even earlier law, the 1906 Pure Food and Drug Act, which required listing ingredients on interstate shipments of foods, as an example of government-mandated “sunlight” to reduce public risks.¹³

   President Franklin D. Roosevelt quoted Brandeis’s words twenty years later when he urged Congress to require new corporate financial disclosure rules after millions of Americans lost their savings in the stock market crash of 1929. The 1933 and 1934 Securities and Exchange Acts ordered publicly traded companies to disclose assets and liabilities at regular intervals and in a standardized format.¹⁴ Corporate financial disclosure as required by those laws, which remains at the core of U.S. securities policy, has become the United States’ most sophisticated – though still imperfect – example of targeted transparency policy.

TRANSPARENCY AS MISSED OPPORTUNITY

However, targeted transparency policies can also do more harm than good. Such policies are always the products of political compromise. When the information from the tug and pull among many interests is incomplete, inaccurate, obsolete, confusing, or distorted, it can contribute to needless injuries or deaths or to large economic losses.

   Four years before successful use of targeted transparency to reduce auto rollovers, Congress tried to enlist the power of information to reduce another serious safety risk – disease outbreaks from contaminated public water supplies. This time Congress failed.

   Drinking water safety became national news in 1993 when a microbe called cryptosporidium infested the drinking water of Milwaukee, Wisconsin, sickening an astounding 400,000 individuals and killing as many as 110 within a matter of weeks. Congress responded in 1996 by demanding that water authorities inform their customers about contaminants in the water supply.¹⁵

   That time, though, Congress crafted a requirement that employed technical terms, produced inaccurate and out-of-date information, failed to link contaminant data to health risks, and did not provide comparability from one community to another. Instead of receiving clear information that was comprehensible at a glance, like the five-star auto rollover rankings, consumers seeking information about the relative safety of their tap water faced the daunting task of interpreting complex documents like that shown in Figure 2. Just at the time when electronic monitoring and the Internet made real-time reporting feasible, water authorities’ lobbying as well as careless planning by policymakers produced partial and hard-to-decipher information that was as much as a year out of date.¹⁶

   As a result, some customers who relied on assurances that tap water was safe actually suffered increased health risks. In a particularly troubling series of incidents, media reports in 2004 revealed that tens of thousands of children in Washington, D.C., Boston, and other big cities were drinking water contaminated with unreported high levels of lead, an especially dangerous toxin that could cause severe neurological damage in children. In the nation’s capital, federal and local officials admitted they had known about the lead contamination for years but had neither informed customers nor taken steps to remedy the problem. Instead, the city’s contaminant reports assured customers, “Your Drinking Water Is Safe.” One reason that risks remained hidden was because contaminant reports did not include information about microbes or toxins that entered water after it left the filtration plant – as it passed through hundreds of miles of old lead pipes.¹⁷

   Drinking water reports represent a missed opportunity with serious consequences. According to the National Centers for Disease Control and Prevention, up to 30 percent of reported disease outbreaks each year can be attributed to problems of public water systems, affecting as many as nine hundred thousand people. In 2005, Stephen L. Johnson, the new administrator of the federal Environmental Protection Agency, estimated that at least 10 percent of Americans regularly drink unhealthy water.¹⁸ All in all, as many as 50 million Americans drink water containing industrial solvents and related chemicals that may have long-term health effects.¹⁹

   In the largest water systems, the mixture and levels of contaminants vary greatly from week to week as weather and waste discharges change. Accurate and current contaminant reporting can be critical for those most vulnerable – the very young, the very old, and people on chemotherapy, suffering from AIDS, or with otherwise compromised immune systems. Such individuals – who together make up roughly 20 percent of the U.S. population – are at special risk from bacteria or toxins in drinking water.

   Meanwhile, the public’s trust in the nation’s water supply continues to erode. A quarter of Americans reported in 1999 that they never drank tap water. Sixty-five percent of those who did drink tap water reported that they drank bottled water or filtered tap water some of the time.²⁰

   Transparency gaps that increase serious risks are common. Some other prominent examples:

•   Millions of investors lost savings and retirement funds in 2001 and 2002 not only because corporate executives at some of the nation’s largest and best-known companies fraudulently withheld information but also because the financial accounting system allowed them to hide – and profit from – information about financial risks in their companies.

•   Millions of people have unknowingly increased their risk of heart disease because nutritional labels have not told consumers when cookies, muffins, and other fast foods contain trans fats, the most dangerous fats on the market. For two decades, scientists have known and warned of trans fat risks.

•   Despite twenty years of alarming evidence that more people in the United States die from medical errors in hospitals than from auto accidents and findings that some institutions are ten times safer than others, hospitals are still not required to disclose mistakes that cause death or serious injury.

•   Five years after the September 11, 2001, terrorist attacks in New York City and Washington, D.C., government officials still rely on a five-color terrorist threat warning system that does not provide the public with needed information for self-protection – leaving information gaps that could cost thousands of lives.

   The cases we have drawn together illustrate both the promise and the perils of a new generation of targeted transparency. By requiring auto rollover ratings, Congress invented a means of communicating complex information in a simple format that helps car buyers compare models and make safe choices. By requiring reports on drinking water safety, Congress settled for a compromise that produced out-of-date and incomplete information that confuses and sometimes misleads customers. Such distorted disclosure not only impairs public health. It also undermines one of democracy’s central tenets – that citizens can trust their government as a source of reliable, timely information.

A REAL-TIME EXPERIMENT

What makes the difference between transparency success and failure and how can its effectiveness be improved? We have written this book to answer these questions.

   We have scrutinized a carefully selected group of transparency policies using a multidisciplinary approach. We have analyzed the effectiveness of fifteen major targeted transparency policies in the United States and three international policies. Out of the universe of policies that fit our definition of targeted transparency, we chose a set of relatively mature cases, distributed across a range of public policy areas, with potentially important consequences, and whose varied effectiveness has been assessed in rigorous empirical studies.

   We reviewed the legislative history and legal requirements of each policy and examined the politics surrounding initial approval and later amendments. We assessed each policy’s regulatory structure and the incentives that structure provided for accurate reporting by disclosers, as well as the kind of information it ultimately provided to users. We also analyzed the decision-making processes of users and disclosers in order to understand their actual responses to newly released information. Finally, we identified the drivers of effectiveness by analyzing hundreds of empirical studies and by interviewing policymakers, scholars, and diverse users and disclosers of information. This approach has allowed us to develop a theory of targeted transparency effectiveness that explains the varying outcomes of existing policies and can provide the basis for future research. Table 1 provides an overview of the eighteen targeted transparency policies that form the analytic core of this book, and the Appendix contains our detailed descriptions of them.