Timothy Hwang ’14: Automating K Street

Even before arriving at Princeton, Timothy Hwang ’14 had an interest in politics, working as a field organizer for Barack Obama in 2008 and serving as a student member of the Board of Education of Montgomery County (Md.) Public Schools when he was 17. At college, Hwang majored in the Woodrow Wilson School but also took courses in computer science. Then, in a Sunnyvale, Calif., Motel 6 during the summer after his junior year, Hwang put those interests together, founding a company with two high school friends that news organizations have said could automate the ways of Washington.

“Lobbying, like baseball, no longer belongs to old-timers and their seasoned intuition: It’s now being refined by computer data and forecasts.” — MIT Technology Review

That company is FiscalNote, founded with Jonathan Chen and Gerald Yao, which brings the principles of big data to Washington, where policies affecting the lives of millions can be won or lost on personal relationships. Hwang originally thought he would take “a more typical Woody Woo-major path” and work on tech-policy issues at the State Department, Federal Communications Commission, or on Capitol Hill — “but then I thought maybe we could do the opposite and take some of that policy expertise and apply it in the technology space,” he says. FiscalNote was designed as a platform that would allow lobbyists and policymakers to track what was happening across federal, state, and local politics — a tool that can quickly sort through and analyze data that, in the past, would have required hundreds of hours of labor. 

Five years later, FiscalNote employs more than 400 people, has 5,000 clients, and has moved from Motel 6 to a sprawling space on Pennsylvania Avenue, complete with a Silicon Valley vibe and a glass-walled conference room called “Princeton.” Last summer the technology company purchased CQ Roll Call, one of the best-known outlets for nonpartisan political journalism in the United States, for $180 million from The Economist Group. It’s one of several companies that recently have brought big data to lobbying. 

“Lobbying, like baseball, no longer belongs to old-timers and their seasoned intuition: It is now being refined by computer data and forecasts,” said a recent article about FiscalNote in MIT Technology Review. Whether that’s good or bad is still being debated. 

What can FiscalNote do? If you’re a subscriber with an interest in a political issue — say, the minimum wage or renewable energy — the platform will pull up pending bills in Congress and state legislatures that deal with that issue. FiscalNote provides an estimate of how likely it is that a bill will pass and populates a “virtual whipboard” that sorts the relevant legislators according to how they are likely to vote (yes, lean yes, toss-up, lean no, or no). It also provides a liberal-to-conservative ideology scale showing where each legislator’s voting record falls on a range of issues from transportation and health to crime and agriculture. 

There’s a tool for automatically analyzing the comments submitted by the public about a proposed regulation, a process of reading and sorting that, done manually, would take many days. It mines bills, comments, and proposed regulations. There’s a search engine for the Twitter feeds of policymakers at the state, federal, and local levels, as well as reports published by the Congressional Research Service and policy documents from countries including Argentina, Canada, Chile, India, and the United Kingdom. (Hwang says a new country is uploaded every three to four weeks.)

The company demonstrated its technical prowess in the spring of 2017, when the Federal Communications Commission opened the public-comment period on its plan to repeal net-neutrality protections. A record 22 million comments (including many submitted by “bots”) poured in. FiscalNote sorted through all of them, organizing them for clients in a way that showed not just whether each commenter was in favor or opposed to the proposal, but for what general reason. 

Hwang sees FiscalNote partly as a tool to help large multinational companies gather the data they need to comply with myriad rules at different levels of government. “If you’re a large food-and-beverage company, then you have these big farms, distribution arms, restaurants, and all of the agriculture and labor and environmental and food-safety regulations,” Hwang explains. “The way our regulatory system is set up [is] complex and byzantine — and then you multiply that across 50 states.” 

Moreover, he sees “big data” analysis as a way to give smaller Washington players access to tools the big players have wielded for years, helping them better target their message and use resources more effectively. Clients include small and nonprofit organizations including the Boys & Girls Clubs of America, the Natural Resources Defense Council, and Meals on Wheels — nonprofits Hwang says could not otherwise afford to analyze comprehensive policy data. 

FiscalNote won’t disclose specific fees, but says they are scaled according to an organization’s size and whether it is a for-profit or nonprofit entity. A nonprofit typically would pay “in the low five figures,” while a large, multinational corporation would pay in the high six figures, according to Chris Lu ’88, deputy secretary of labor during the Obama administration, who now works at FiscalNote. Meanwhile, a 2013 New York Times article about lobbying expenses for small businesses estimated that hiring government-affairs professionals to represent them could run $5,000 to $20,000 per month.

Hwang sees “big data” as a way to give smaller Washington players access to tools the big players have wielded for years.

But others are skeptical of the argument that big data will give a meaningful voice to groups that have not been able to afford access to power. “These kinds of tools usually become weapons of powerful interests and organized groups because those groups already have an apparatus up and running to use that information,” says Tim LaPira, a political science professor at James Madison University who previously worked at the Center for Responsive Politics. 

Benjamin Waterhouse ’00, a history professor at the University of North Carolina at Chapel Hill who wrote Lobbying America: The Politics of Business from Nixon to NAFTA, says just giving smaller organizations access to more information may not be enough to amplify their voices in policy debates. “There’s a tendency to over-valorize the power of small businesses and groups, and giving them more data may be part of imagining that those smaller voices can have a serious impact, but the power centers in lobbying have always been the biggest, richest organizations,” he says. 

Indeed, Hwang insists that FiscalNote aims to help lobbyists with research and tracking, not to replace them with fully automated software counterparts. Princeton computer science professor Edward Felten, a former U.S. deputy chief technology officer who serves on a task force looking at how the labor market can adjust to advances in artificial intelligence and technology, seems to agree that lobbyists need not worry about unemployment. “To me, [FiscalNote] sounds less like a robot lobbyist and more like a robot research intern,” he says. “A government-relations person is trying to understand what policymakers are thinking; they’re looking at legislation and other proposals and trying to figure out what’s going to happen — and that’s going to require human judgment.”

Notes Waterhouse: “Even in an age of big data, lobbying still operates the same way it always has — based on personal connections.”