Meet Edward Tufte, the graphics guru to the power elite who is revolutionizing how we see data.
Tufte may be one of the few theorists—design or otherwise—lauded in both the Bush and Obama administrations. After their first meeting on Pennsylvania Avenue, Earl Devaney sent Tufte’s resume to the White House, leading to his appointment last March as an official adviser to the Recovery Transparency and Accountability Board. But his influence in the administration has reverberated beyond that office. Peter Orszag, who left his post as head of Obama’s Office of Management and Budget last year and then took a job as vice chairman of global banking at Citigroup, spoke to me recently in his office on the thirty-ninth floor of the bank’s headquarters in lower Manhattan. He was first given a copy of Tufte’s Visual Display of Quantitative Information as a college graduation present. As Orszag explained, the challenge of designing graphics to represent the nearly infinite stream of data coming out of government offices is to be compelling but not to oversimplify. The description sounded like many of the maxims I heard on tour with Tufte. Indeed, at the end of our conversation, Orszag told me that “Tufte’s spirit” has “been taken to heart by an entire generation of data analysts.”
Before his work with Devaney, Tufte’s most public engagement with the federal government came in the form of a scathing crusade against its endemic use of Microsoft PowerPoint. In January 2003, the space shuttle Columbia took off from Kennedy Space Center. Seconds into its flight, a piece of foam insulation broke off from the shuttle’s fuel tank and punctured its left wing. For days, as the shuttle orbited the earth, engineers at NASA and Boeing debated how serious the hole might be. Much of this communication was done in the form of PowerPoint, a software program that Tufte says is constricting and obfuscating and “turns information into a sales pitch.”
After reviewing several PowerPoint presentations, NASA officials decided to proceed with the shuttle’s return flight. On February 1, Columbia burned up as it reentered the earth’s atmosphere, killing all seven crewmembers on board. Tufte has called PowerPoint a “coconspirator” in the shuttle disaster.
Tufte dissected NASA’s PowerPoint slides on his Web site, showing that the program didn’t allow engineers to write in scientific notation and replaced complex quantitative measurement with imprecise words like “significant.” He then published a twenty-eight-page essay called “The Cognitive Style of PowerPoint,” in which he analyzed hundreds of existing PowerPoint slides and showed that the statistical graphics used in PowerPoint presentations show an average of twelve numbers each, which, in Tufte’s analysis, ranks it below every major world publication except for Pravda. The low information density of PowerPoint is “approaching dementia,” he wrote.
Later that year, in August 2003, the Columbia Accident Investigation Board issued its own report on the shuttle accident. On page 191, in a section called “Engineering By Viewgraphs,” the board cited Tufte’s research, and wrote that it “views the endemic use of PowerPoint slides” as “an illustration of the problematic methods of technical communication at NASA.”
While PowerPoint still takes a rather severe lashing at Tufte’s courses, he says he has largely moved on from the debate. “It’s a public and corporate disaster for Microsoft,” he said. “But it’s a trivial issue to me.” The general backlash against PowerPoint has only grown, however. In 2009, Thomas X. Hammes, a retired Marine Corps colonel who is now a fellow at the National Defense University, wrote a widely circulated essay in Armed Forces Journal that argues that PowerPoint has had a “toxic effect” on the culture of debate and discussion within the U.S. military, claiming that “senior decision-makers are making more decisions with less preparation and less time for thought.”
As Hammes explains it, the reliance on PowerPoint often means that battle orders are rendered in incomplete, often unclear sentences and maps are squashed and stripped of meaningful detail, leaving essential battlefield questions of geography dangerously unclear. The details are classified, but Hammes told me that he has seen war plans for the Korean peninsula prepared in PowerPoint in which massive terrain issues were completely glossed over. On the whole, Hammes told me, the rise of PowerPoint in the military has made the decision-making process less intellectually active. And Tufte, he added, “is the master on this whole thing.”
The spread of technology over recent decades has fostered a comfort with data, and an appetite for more of it, in an ever-increasing portion of the general public. This cultural shift was especially visible in the dense data displays that appeared in newspapers, and especially online, in the run-up to the 2008 presidential election. Here, too, Tufte’s influence was evident. Nate Silver, who runs the political Web site FiveThirtyEight, now part of the New York Times, uses many of Tufte’s maxims in the site’s design. Silver told me that he tries to keep the “data-ink” ratio of his current site very high, meaning most of the pixels on the screen show actual numbers or data points; he also thinks of the site’s design in terms of “small multiples,” another Tufte neologism that refers to a series of related numbers that reveal subtle differences over time. “Tufte treats data like good writing,” he said. “You have a certain thought—how clearly and beautifully are you conveying it?”
Good design, then, is not about making dull numbers somehow become magically exhilarating, it is about picking the right numbers in the first place. “It’s about data that matters to you,” said Dona Wong, who was a student of Tufte’s at Yale in the 1980s and later the graphics director at the Wall Street Journal.
This idea is especially relevant for how the federal government measures and displays the impact of the stimulus. What are the most informative comparisons? And what, for lack of a better word, is simply chartjunk? Tufte first suggested that the recovery board rethink how the projects are ranked. Until now, they have been sorted by location, with the idea that any person could look up projects in his or her county or zip code. “But what’s it matter where the project is? We already know where Texas and California are,” Tufte said. Geographic space helps little in this context. “Always order things substantively,” Tufte explained. He would like to bring in all other kinds of data sources to add context to the stimulus and track its actual effects. Soon, if Tufte has his way, Recovery. gov will display analytical graphics that measure funded projects alongside metrics like infant mortality, traffic flows, and education. After all, as he told me, “understanding doesn’t come from project descriptions.”
Someday, the Web site might also use what has proven to be one of Tufte’s more popular inventions. In his 2007 book Beautiful Evidence, Tufte introduced what he called “sparklines,” numerically dense, word-size graphics that show variation over time. They have since appeared on the financial pages of Yahoo! and the sports section of the New York Times. (As an example, the fluctuations of the Dow Jones Industrial Average over the course of February and March look like this: . The dramatic dip in the figure represents the March 16th panic over the nuclear disaster in Japan.)
Feed the Political AnimalDonate
Washington Monthly depends on donations from readers like you.