Since 2000, homeownership fell in 90 out of the top 100 largest metros in the U.S., according to the latest report from Trulia.

Nationally, 72.4% of all larger ZIP codes, or those with at least 1,000 occupied housing units, saw a decrease in their homeownership rate from 2000 to 2016, according to the report.

Over the past few years since the turn of the millennium, several cities have shown dramatic shifts from homeownership to renting such as Cape Coral, Florida; Las Vegas and Phoenix. In one Phoenix neighborhood, for example, homeownership plummeted from 92.6% in 2000 to 50.5% in 2016 after the ZIP code 85305 gained nearly 2,000 new apartment units.

Other metros, especially in the Northeast, countered the national trend with an increase in their homeownership rate since 2000. One neighborhood in Atlanta, ZIP code 30313, saw the greatest gain and, as its median household more than tripled, homeownership increased from 13.1% in 2000 to 32.5% in 2016.

In some areas, such as Houston and New York City, the homeownership rate has remained unchanged, yet several significant shifts have occurred within the city itself, according to the report.

The national homeownership rate rose from 67.1% in 2000 to 69.2% in 2004. However, after the Great Recession, that rate came crashing back down to 62.9% in mid-2016.

“Taking a closer look at some of the ZIP codes that have swung the furthest toward more renters among the 100 largest metros, we see that none of them are in any of the northeast metros, and most are in the southwest, Texas, or Florida,” the report stated.

The chart below shows which U.S. metros held the most ZIP codes with increasing homeownership rates from 2000 to 2016.

Trulia explained in its report that metros with higher rates of homeownership increases also saw higher construction rates among single-family homes and townhomes, rather than multifamily units. Also, in some metros, a change in median household income, for better or worse, helps explain the direction of homeownership in the area.

“Ultimately, our research found that neighborhoods that swung to renters from owners or vice versa were influenced by three factors: changes in household income, the type of new construction, or lack of it, in the area and the housing crisis which ultimately displaced huge swaths of the population,” Trulia stated in its report. “While it’s impossible to predict what will happen next, as incomes rise or fall, new rentals, or homes are built and instability – these factors always will have an outsized influence in whether a neighborhood is full of homeowners or renters.”