When we look back on the current housing downturn a decade from now, we’ll remember many things: The foreclosures, the sinking home values, the consolidating households and the tough decisions many families had to make. But, more than in any other housing recession, one statistic will stick in our minds. Negative equity—the percentage of homeowners who owe more on their mortgage than their home is worth—is both the symptom of our housing excesses, and the disease that has kept the housing market on its knees.

Last week, the Wall Street Journal ran a story dissecting the methodologies behind several prominent negative equity estimates—Zillow’s included. The writer, the WSJ’s “Numbers Guy,” Carl Bialik, did a great job of explaining how all the different companies go about calculating negative equity.

Here at Zillow, we calculate negative equity on a household level by looking at the total amount of mortgage(s) and loans taken out at the time of purchase, and comparing that to our current valuation of the home. We calculate this quarterly, and in Q3 2010, found that 23.2% of owners of single-family homes with mortgages were underwater. That was up slightly from 22.5% the previous quarter. Bialik also looked at the negative equity statistic created by CoreLogic, which uses similar methodology.

Despite the thorough analysis, we had a couple issues with the story and some of the comments within. Below, we summarize the two areas of concern and address them each.

1. Just because a homeowner is in negative equity does not necessarily mean they will default on their mortgage. Therefore, estimations of negative equity make the problem seem bigger than it is. Bialik interviewed Kenneth Rosen, the chair of the Fisher Center for Real Estate & Urban Economics at the University of California, Berkeley, who said, “Everyone likes to get headlines, so they tend to overstate problems like this, without doing ground-level research.”

First, Zillow’s negative equity numbers represent nothing more than the percentage of homeowners who are in negative equity. It’s true that not every person who is underwater on their mortgage will default. In fact, we’ve long made the point that, for most people, being in negative equity is a lot like experiencing an unrealized loss in the stock market—it probably won’t not matter until it’s time to sell.

However, that doesn’t mean it’s ok to dismiss the statistic. While negative equity may prompt conscious, proactive decisions—like strategic defaults—in only those who are extremely underwater (see this research, among the many, for a discussion of this effect), negative equity at all levels can have adverse effects. Any homeowner who is in negative equity and who experiences a financial shock, like a job loss or divorce, will find their options limited (see, for example, the Background section of this research from the Boston Fed). To sell a home while underwater means the seller must pony up the difference between the mortgage and the current value, or negotiate a short sale with their bank. Therefore, negative equity does open a lot of homeowners up to risk they would not otherwise encounter, and it’s important to understand the total universe of homeowners who have this increased risk.

2. The statistic that one-in-four homeowners is underwater “is calculated using assumptions that inflate the number of underwater homes.” Bialik concluded that, because Zillow and other firms that calculate negative equity use valuations, each which has some margin of error, and because the number of homeowners just barely in positive equity is somewhat greater than the number who are just barely in negative equity, the estimation error will tend to incorrectly place more people in negative equity (who are, in fact, in positive equity) than is offset by incorrectly placing people in positive equity (when, in fact, they are in negative equity). This subtle point has the possibility of producing an upward bias in the negative equity statistic.

This is a fair point and one that I conceded. The question, however, is one of magnitude. While I agree that the shape of the distribution of loan-to-value around the critical 100% threshold (separating negative and positive equity) can produce subtle biases, I suspect these biases are relatively minor relative to the absolute number. That is, I don’t believe that accounting for this possible bias would change the negative equity estimate from 23.2% to 15%, for example.

Moreover, I would suspect that this effect is not nearly as important as the fact that the standard approach to computing negative equity does not incorporate the costs of actually selling a property. Considering these costs, which can be six to eight percent of the price of the home and must be subtracted from the sale proceeds before a seller can take any equity that might remain, would significantly increase the number of homeowners who are effectively in negative equity should the need arise for them to sell their home.

Finally, I have a general concern with the message that some readers may take away from the article. Some will find it a fascinating deep dive into a very important metric for the housing market—an exploration of the complexities of computing data—which is, I believe, the spirit with which Bialik approached the topic. Others, I fear, will come away with a certain data nihilism, believing that there is no truth or, if there is, we cannot know it. To discount the negative equity statistic altogether is to suggest either that it doesn’t matter or that there is no way of understanding truth in the market. But it does matter—many Americans are underwater on their mortgage, this fact is translating into foreclosures and this dynamic is vitally important to understanding the current real estate market—and the truth is out there. Zillow, CoreLogic and Moody’s have put a lot of thinking and resources into estimating negative equity as accurately as possible. In addition, all responsible parties have published negative equity statistics that cluster around one another, which strongly suggests they are close to the truth. While the estimates may be imperfect, it’s wrong to conclude there is no truth and that understanding the full scope of the country’s negative equity problem doesn’t matter.