That Data You Depend on May Not Say What You Think

Data-driven decisions are critical in business. Have the right information and you’re well on the way to a smart decision.

Government data can be a gold mine for executives. In real estate, one of the prime sources is the Census Bureau. Forget for a moment the population counts, economic projections, and demographics. Here’s where you get the data on housing starts, sales, and construction for direct real estate information. Other important data points include retail and food service activity.

“Rental vacancy rate trendlines can materially inform a multifamily investor’s underwriting decisions related to occupancy forecasts,” Josh Fischer, a principal and director of acquisitions at real estate private equity investment firm Birgo Capital, tells GlobeSt.com. “If a given property has a current occupancy rate of 98%, but the vacancy rate of the larger metropolitan area has been steadily increasing by 2% a year, said investor would be wise to consider how this macro dynamic is likely to affect the specific acquisition they are considering.”

But using data has risks, particularly when what you receive isn’t the full information. Many reports on Census data do not include the statistical limitations and uncertainties that may turn a decision from a no-brainer to a head-scratcher.

Take the Census Bureau’s 2021 first quarter residential vacancies and homeownership statistics released Tuesday as an example. National vacancy rates in the first quarter of 2021 were 6.8%, compared to 6.6% in the same period of 2020. Homeownership rates were 65.6% in 2021 Q1 and 65.3% in 2020.

And yet, according to the Census Bureau, the differences between the two years were not statistically significant. That means from a mathematical point of view, the difference was so close that they might have been a limitation of measurement. The margin of error in the 0.2 percentage point difference between 2020 and 2021 was 0.3, meaning that the relative positions could be backwards.

Or take the March 2021 monthly new residential sales. At a seasonally adjusted annual rate of 1,021,000, they were 20.7% above the revised February rate of 846,000. Except, the new numbers were plus or minus 23.7%, which means they could have been 43.7% above or 3% below. Because a change of zero lies in that range, called the confidence interval, “there is insufficient statistical evidence to conclude that the actual change is different from zero,” according to the Census Bureau. There may have been a small drop, no change, or a big jump. There is no way to know.

Advanced retail sales for March offers another example. Many outlets trumpeted the 9.8% growth. But, again, there was a caveat: “The 90 percent confidence interval includes zero. There is insufficient statistical evidence to conclude that the actual change is different from zero.”

Unfortunately, most media outlets in their reporting don’t mention these uncertainties, which can have unintended impacts on decisions.

Jeff Shipwash, founder of Shipwash Properties, tells GlobeSt.com that he uses “census sales data in my real estate business to adjust which markets I want to target.” When asked whether he knew about the variations, he says, “I actually did not know that! That is great information to know moving forward. This could impact our timelines moving forward on marketing targeted areas.”

“If you’re going to be in the commercial real estate space, you have to be a demographic guru,” Michael P. Feldman, CEO of Choice NY Companies, tells GlobeSt.com. But he finds that statistical information is often badly reported. “Within 24 hours, I saw one article that said multifamily rents are increasing drastically and another said they’re not moving,” he says. “It can’t be both.”

But, if you aren’t getting accurate reports of what the Census Bureau says, oh, yes it can.