In 1995, when Alan Ziobrowski was an associate professor of finance at Lander University, in South Carolina, he found himself at home one night watching “one of those 60 Minutes–type shows.” That evening’s story caught his interest: Gregory Boller, a professor of marketing at the University of Memphis, had found some striking coincidences in which members of Congress, between 1990 and 1995, bought or sold stock in companies that could be affected by ongoing government activity.
According to Boller’s study, which Mother Jones also covered, Senator Lloyd Bentsen (D–Texas) had bought stock in a dairy processor and sold it 10 months later, days before the Justice Department began investigating the company for rigging bids to sell milk in public schools. Senator Bob Dole (R–Kansas) had purchased stock in Automatic Data Processing four days before President George H. W. Bush signed a law with new rules for military data processing. Representative Newt Gingrich (R–Georgia) bought Boeing stock just before he helped kill amendments that would have cut funding for the International Space Station—an outcome that helped Boeing secure a contract.
It all sounded very damning. And yet to Ziobrowski, Boller’s work didn’t seem completely fair. The examples were cherry-picked and could, in fact, have been mere coincidences. How many times, for example, had congressmen sold Lockheed Martin or AT&T right before passing laws that benefited the companies? To know whether members of Congress were turning insider knowledge into personal financial gain, you needed to look at all their trades—those made by the small fry as well as the presidential contenders, the losers as well as the winners. The real question, he thought, was whether a portfolio made up of stocks held by members of Congress would significantly outpace the market.
To find the answer, Ziobrowski recruited three other professors, including his wife, Brigitte, who oversaw the tedious process of actually turning the entries on the opaque disclosure forms into usable data. While the House forms were available in libraries, getting the records from Senate offices proved especially difficult. “We finally had to buy many of them at 20 cents a page,” Ziobrowski told me. And even after they got the forms, “it took years to go through and track all of those transactions,” Ziobrowski says.
Their results were necessarily approximate. Some members left office before selling shares; others would sell some, but not all, of their stock. “And then there are others who just don’t bother to tell you when they sell,” Ziobrowski told me. “You’ll see the portfolio change, but you don’t know why. There’s no watchdog, there’s no one who audits to check that they’re accurate.”
Even so, the professors eventually had a data set comprising all known senatorial transactions between 1993 and 1998. And what they found shocked Ziobrowski.
“Most of the time, you do studies like this, and you end up concluding that there are no abnormal returns. Call us naive, but the part that bothered me most about Boller’s work is that it suggests that they’re doing something sneaky, but it didn’t actually show that they were doing anything sneaky.” He chuckled. “None of us were betting our tenure on the results of this study.”
But when they’d completed their analysis, it looked like they—and at least a few members of Congress—had hit the jackpot. Their analysis of the Senate returns over the six-year period, published in the Journal of Financial and Quantitative Analysis in 2004, showed that the Senate portfolio outperformed the market by approximately 12 percent a year. In April of this year, the team published a follow-up analysis of the House in Business and Politics, which showed that House members on average outperformed the market by a smaller but still impressive figure—roughly 6 percent a year.
These numbers are bigger than they sound. If you took $100 and invested it at 6 percent over a 40-year career, when you retired, that $100 would have increased almost tenfold … while if your portfolio averaged a steady 12 percent, your original $100 would have turned into more than $8,000. If you’ve looked at your 401(k) recently, you know most people don’t get such returns. And that was just the extra profits they made, over and above the 20 percent annual return that even a blind monkey with a dartboard and an E*Trade account was making back in the late 1990s. A study of corporate insiders, who presumably have lots of information about their firm’s performance, showed that their purchases earned abnormal returns of only about 6 percent a year. Senators seemed to be the greatest stock pickers since Warren Buffett.
But of course, Warren Buffett spends most of his days locked in his office in Omaha, pondering his investments. Senators had to get their stock deals done between rubber-chicken dinners and grip-and-grins at the state fair. Which made it hard to escape the conclusion that they were doing something a little worse than sneaky.
Or were they? A few years ago, after the Center for Responsive Politics made congressional disclosure forms available in an easily searchable data set, two graduate students in Harvard’s political-science department decided to revisit the question for the new millennium. Andrew Eggers and Jens Hainmueller, now assistant professors at the London School of Economics and MIT, respectively, had become interested in the emerging literature on whether politicians benefit financially from holding office, and they were just finishing their first paper, which used probate data to look at that question in the United Kingdom. (Answer: yes, at least if they were Tories in the House of Commons between 1950 and 1970.)
They eagerly attacked the U.S. data. Both of them got a big surprise. Their data, which covered 2004 through 2008, didn’t show Congress outperforming the market by 12 percent. In fact, they didn’t show it outperforming the market at all. For the five years they studied, Congress actually underperformed the market by 2 to 3 percent annually. On average, Congress did worse than an index fund, and about as well as your average stock-picking granny. If Congress was indeed trading on inside information in the late 1990s, it seemed to have stopped.
But, if all the data are right, why would congressional stock-picking have changed so much? That’s “the one fact that we cannot 100 percent nail down,” says Hainmueller ruefully, especially since Ziobrowski et al. haven’t released their painfully assembled data set. One of the papers could be wrong, of course, but it’s hard to adjudicate that when one group hasn’t released its data and the other’s paper hasn’t yet been published in a peer-reviewed journal—and anyway, neither group is openly disputing the other’s results. Still, Eggers and Hainmueller’s difficulty in explaining the difference may make it harder for them to get their own results published.
One possibility is that insider trading has gotten harder. As a risk-arbitrage trader at Goldman Sachs, former Treasury Secretary Robert Rubin made a living exploiting persistent anomalies between, say, the stock price of a company that was being acquired, and the price that the buyer was offering; now hedge funds are in a technological arms race to gain advantages that last for only milliseconds. Meanwhile, a booming political-intelligence industry scrutinizes Congress like a flock of half-starved vultures. Congressional information advantages may not persist long enough to let members profit from them.