An app that could change your vote: Nikita Bier at TEDxBoston

So back in 2011, I was a student at UC Berkeley studying public policy and protests surrounding tuition hikes were gripping the school. However, many years ago, the university had adopted a policy whereby only a handful of high income students pay more when tuition goes up. In fact, the additional revenue collected from those students actually gets shifted down to most other students in the form of bigger financial aid packages. I created this visual model of how the policy works and we found that most of the people out there protesting were primarily low and middle income students who would actually benefit from the hikes. And this was circulated all over California. The protests diminished and the president of the university sent me this letter explaining his trouble in communicating the policies. It was clear to me that our notions around tuition hikes is that it will limit access to disadvantaged groups. However, it's often the case that rhetoric will supersede logic when our public institutions are so complex that the ordinary citizen doesn't have the hours and the day to understand them. And the consequence of all this is that we as citizens unknowingly advocate for things against our own interests or the interests of those we care about. In this creates unnecessary gridlock. It distorts our voice into government, and then government can't perform its job of serving people. So while the president of my university just wanted to make college more affordable, just like the protesters had wanted, he faced immense barriers for really no reason at all. So I started wondering if this problem existed outside the walls of my university. Were people's beliefs consistent with how public policy would impact them? So I started a project and we looked at how the Obama and Romney plans would impact different zip codes in the United States. We took their tax and spending proposals, we turned them into algorithms, and we simulated the impacts against a bunch of tax records and census records. And we discovered something very big and that's most Americans vote against their own financial self interests. Now I know it's a common belief that many red states do this because of social issues. I mean, Thomas Frank wrote a best seller about it called what's the Matter with Kansas? Where he argued that the Republican Party leverages non economic issues to win votes in middle America. But this non economic voting is much more pervasive than we think. Some of the most liberal parts of the country happen to be some of the biggest benefactors of Republican policy. Here's San Francisco. The red areas are benefited by Romney's plan. Here's New York. I started to wonder if this was just an American thing, was this happening in other countries? So we took a look at other countries in the western world. And we found that the US actually has the lowest rate of class based voting in the entire western world. That is to say, we simply don't consider our personal finances when we go to vote. So the ironic thing, our common thread as Americans is our shared ignorance around public policy. But a lot of this has to do with our country's diversity. We get divisive about other issues like culture and religion, while in more ethnically homogenous countries, the biggest divisions are social class. So their political parties tend to form around those things. But I had a feeling that what was happening in America wasn't deliberate. I felt that most of us at our core are moderates. Most of us aren't going to forego the livelihoods of our friends and family for ideology. Most of us simply want to live better lives and in better neighborhoods. And that's not to say that self interest should be our only criteria when voting, but we should at least know the impacts of our options. So while I was still a student at Berkeley, I started building a web application with my friend Jeremy. And it was a tool that could show how much you would pay in taxes and what benefits you would receive under the Obama and Romney plans based on your own demographic information. It took a few months to develop. We didn't know what to expect really. But we put it online one night and we emailed a few blogs and I went to bed. And the next morning my phone started buzzing with calls and emails from journalists. Even though it wasn't a cat video, it had gone vin. We started getting interviews from all over the world. By election day, it was used over 4 million times. Somehow it really struck a chord with Americans, but it also agitated many people. Here's some emails from our users. When information contradicts our preconceived notions, we often reject the validity of the source or we retreat to higher moral ground. Here's a journalist who refused to write about us and criticized our mission. But all of this, it didn't bother me. All that mattered to me was if we were helping Americans form sound opinions around public policy. And fortunately, a team of professors at Berkeley did a controlled study on our site and they found that 6% of our users shifted their support to the plan that benefited them more. That was shocking because that was enough to change the outcome of an American election. And this call to rational thinking, this influence started attracting the attention of governments from all over the world. They started emailing us, asking us to build an analog for them. They told us that they have no way of communicating the pros and cons of policy in a way the ordinary citizen could understand. They can't even do the right thing anymore because of how media and politics would distort proposals. So we started thinking, was there a way to scale what we did for the election to every policy issue so that could be the universal form of communicating this information. So I put together a bigger team and, and we embarked on a not so trivial task and that was developing a model of the US economy. We spent about a year merging data sets from places like the IRS and Census. It culminated in over 200 billion data points on the US population. And don't worry, they're anonymized. I won't be selling them to the nsa. And what this allowed us to do was measure the impact of any change to a parameter of economic policy down to the individual level. So this became our new company and it's called outline.com and our mission is to help governments improve the well being of Americans. It's really simple. Now I know you guys are a smart audience and you're probably wondering how does he do it? What do you mean? Model. So I'm just going to do a little quick lesson. So there's two types of economic modeling. There's partial equilibrium and dynamic modeling. Partial equilibrium is the kind we do and that measures the first order impacts of a decision. The the direct transfer of resources from household to government. Dynamic modeling is the kind we abstain from. It measures the second order effects and beyond how effects feed into each other. For instance, if you spend more on education, sometimes it lowers crime. We realized that the impacts of those types of second order effects are very uncertain, but also highly politicized. So we abstain from that. And we realize that most people don't even know the first order impacts. So we can't even begin to have a conversation about what would happen subsequent to them. So now I'll stop talking and I'll jump into a demo of how it actually works. If we could switch over. So this is for the Commonwealth of Massachusetts. This is a tool we built for Deval Patrick's administration. It was a pilot. So now I'm going to perform a tax cut. Here you can see every zip code in Massachusetts. So I'll zoom in to where we are right now so we can get a little closer. So these are all the taxes that I can cut. So I'm going to move this slider over and see what happens. So I'm going to cut it maybe down to there and in Real time, it'll actually calculate it against millions of tax records on the back end. You also see that I bankrupted the state. I'm sorry, but let's say I want to make up for that tax cut with a cut to education. So. So now we can look at how it impacts different income levels from rich to poor. So go over to the spending tab, and maybe we'll cut education by a few billion dollars. And then in real time, you'll see that the bottom two income quintiles were impacted most. So now I could take this set of policies and I could actually find out how it impacts a particular demographic. So, for instance, if this person was single, they drive, they make $50,000 a year, maybe they trade stocks and they made $1,000 in capital gains and they have two kids. So each of the policies is itemized in this scenario. This person would gain $625 from this particular set of policies. You can see education right there. So we showed this to governments, and they were shocked by the speed that we could do this analysis. What used to take them months to create and assess a single policy, they could now come to our site and build out the infrastructure of government in real time. And not only that, it could be made digestible completely to the public. And so it shifted from a communication medium to a policy making tool. And most importantly, it was open to the public. So anyone can propose ideas now with the same understanding of the impacts as an economist. So we're now working with a handful of governments across the country, and we're expecting around 10% of American economic policy will be made with the aid of outline. But I want to take a step back and sort of explain what the big picture here is. So, in the 15th century, printing technology allowed for ideas to come from a new source, and that was the public, and it shifted from priests and monarchs. And that facilitated a revolution in scientific, social, and religious thought. Design and data visualization are lowering the barriers for us to influence and engage with public policy. And this will allow all of us, beyond economists and policymakers, to have a sophisticated and constructive conversation about how our public sector should be organized. Thank you.