Steve Ballmer is the former chief executive of Microsoft, and the current owner of the Los Angeles Clippers. This is the second part of an interview with him about his website, USAFacts, which organizes data to provide information about the U.S. population, government finance and government policy. The first part is here. Both parts of the interview have been lightly edited.

HF — You’ve said in the past that while you have pretty open views on politics, you do think that government budgets ought to balance like a business. Will better data and a more balance-sheet-focused approach help push toward this?

SB — Let me separate the chicken from the egg. The number one thing I advocate for is the data. I am a pro-facts partisan. I will push for more accurate, better reconciled, more timely and consistently kept data. That agenda supersedes all others, because there is no integrity if that’s not there. Below that, I have many opinions as most people do on many things. But there are two things I am explicitly partisan on. Number one, I think that every kid born in America ought to have a shot at the American Dream. Many kids don’t, and my wife, Connie, and I will advocate for those things. We’ll use the same data, and be very careful about keeping the data accurate and not in any way influencing it from our perspective.

The second thing is that I do believe that over time budgets have to balance. There are reasonable people who disagree on that. Some people think it’s okay to run a deficit every year, but I don’t understand that. It’s counterintuitive — I’ve studied enough economics to know there’s a case and I am open about that being a bias I have. Are there many ways to get there? Sure. You can raise taxes, you can reform spending, or reform savings plans such as Social Security and Medicare. And on that, I am silent. People have to figure that out.

HF — If you’re a partisan for facts and data, you must be aware of the current political disputes over which data to use for which purposes. Do you think that your approach can help limit some of the fallout from these disputes?

SB — Some of this is about what has happened, but most of it is about forecasting. We’re not in the forecast business. Lots of people can make different forecasts based on the history and the policy. The CBO [Congressional Budget Office] does it — they’ve been a respected source for a number of years. Early in my retirement, I visited the CBO and sat down with Doug Elmendorf and his staff and was impressed by them and what they do. Obviously, the OMB [the president’s Office of Management and Budget] has staff to do these things, and there are various others such as the JCT [Joint Committee on Taxation]. But they’re forecasters. Forecasters have an easy time disagreeing with each other. When it comes to the past, the biggest issue is reconciling the data that exists, making it more timely, and getting into a position where we’re not revising it continuously. There is a significant amount of data that changes after the fact — which is published and then changed. We change it on our site — that probably disturbs many people and makes them worry about the integrity of the data that is continually being revised. Some of these things are done with statistical sampling, obviously. They’re not all spit out of a computer system. I think that requires more care and attention.

HF — Have you seen people who want to get into a fight or disagreement taking up your data to argue with the other side?

SB — I see that even in my own life. My wife and I get started on something — I say no, this only gets x percent of the federal budget, and it is not as significant a percentage as you think or say. She’ll grumble back, anchored again in the numbers. She doesn’t dislike numbers but it’s not where her mind goes first. Once we get grounded, we get back to discussing the real issue. We may agree, we may disagree.

For example, take housing — Section 8 and other housing programs. Should we spend a lot more money or not? Should it be a uniform benefit offered to people or not? These are good questions. My wife and I both read this book, “Evicted” [by Matthew Desmond]. It’s a very good book. People can disagree about what to do, but the first thing to know is how much money is being spent, how many beneficiaries there are, how do we think about that, and then move somewhere.

HF — It seems to me that your project works well with education — providing resources for colleges and schools. Is this something that you’re thinking about seriously?

SB — We’ve actually been pushed on it, more than I expected. I hadn’t even thought about it, frankly, when we first launched. We had to have people push us on it. I think we will wind up trying to work together with a number of people who want to do an educational curriculum from lower schools on up. Do we need to be involved in the creation of that material or even partner with them? No, you can come to our site and do it if you want to. An API might help, but if you’re trying to write a curriculum for 11th-grade civics, you might not even need an API. Just look things up.

HF — Those are my questions. Is there anything final that you’d like to say?

SB — Yes. We’ve learned from the poll we did that Americans do really want the facts. They don’t want anecdotes. They really do worry about the information they get being biased today. Interestingly, 75 percent of Americans think civics education in the country is not very good. People think that if people have different beliefs that they often have different facts, that many times they are not operating on the same fact-base. People do trust politicians more who use real numbers and real data. People do believe that one another change their beliefs on the basis of information. Nobody trusts much of the data sources, whether it’s government websites (shockingly, 40 percent of people have visited a government website — that may include enrollment in government benefits programs as well) or the media. The least respected is social media, and yet people use it the most, particularly younger people.

There’s a real issue here. I used to have this issue at Microsoft, where people would be railing at each other about something. Someone would say, “I believe we need to do this for our customer.” Or “I believe we ought to do that.” And then you’d take a look at it. Let’s say there was a pricing issue — people would be talking about the principle and I’d say, “How much are we talking about in the price difference? How little is the difference if you look at it in numbers instead of the principles?” We’re not asking people to abandon their principles. But in a business situation, if my principle leads me to conclude that we should price our product at $45 and your principle says we should price our product at $50, there is probably a way to get this in the middle. If my methodology says “charge $20” and your methodology says “charge $150,” we probably can’t. But at least we should know where we both stand, and properly recognize the chasms we might hit.

Take anti-poverty programs right now. The president’s budget certainly has an edge on the knife on anti-poverty programs. Some of them were pretty small to start with, and you can discuss whether the problem is the money or the principle. Some of them were very large. You can still talk about whether it is the money or the principle. I’m not arguing in the slightest for or against the president’s budget, but some partisans will say “go after everything,” and some will say “go after nothing,” and the truth is that you should look at the numbers and figure out how far apart we really are.

HF — So the short version of your argument is that numbers have two good consequences — they allow you to figure out agreement where you mightn’t see it otherwise, and where there is disagreement, they make it clearer and more useful.

SB — Exactly.

This article is one in a series supported by the MacArthur Foundation Research Network on Opening Governance that seeks to work collaboratively to increase our understanding of how to design more effective and legitimate democratic institutions using new technologies and new methods. Neither the MacArthur Foundation nor the Network is responsible for the article’s specific content. Other posts in the series can be found here.