In this episode of The REconomy Podcast™ from First American, Chief Economist Mark Fleming and Deputy Chief Economist Odeta Kushi continue season two of The REconomy Summer School series with an in-depth look at the supply side of the housing market, explaining the long-term structural shortage in the nation’s supply of housing and the challenges limiting builders’ efforts to build more homes in the places people want to live.
Don’t miss a single REconomy episode, subscribe today.
Listen to the REconomy Podcast™ Episode 120:
“More supply is the long-run answer to housing affordability. More homes where people want to live means more choices and helps moderate price and rent growth.” – Odeta Kushi, Deputy Chief Economist
Transcript:
Odeta Kushi - Hello and welcome to episode 120 of the REconomy Podcast, our third episode of this year's Summer School series where we discuss economic issues that impact real estate, housing, and affordability. I am Odeta Kushi, Deputy Chief Economist at First American, and here with me is Mark Fleming, Chief Economist at First American. Hey Mark, so we spent the first two summer school classes covering demand. First, headship rates, and then what's driving and delaying homeownership.
Mark Fleming - Right, and today we're going to switch over to supply. Or as I like to say, today we learn how we built this city, housing edition. Getting into it early. Sounds like there will be a lot of construction puns in this episode.
Odeta Kushi - A Starship ‘80s reference so early in class. Well done. You really nailed that. But back to class. Let's start with the basics. When we talk about housing supply, what are we actually counting?
Mark Fleming - It's tempting to think of it as simply all the houses that exist, but it's a little more nuanced. We look at two big categories. Flow measures like new housing starts or completions, which show how much new stock is being added. And stock measures, which capture the total number of housing units at any point in time.
Odeta Kushi - And not every unit in the stock is actually available to live in. Some homes are vacant because they're under renovation or uninhabitable. So, the headline numbers can really hide what’s actually usable. We'll talk about that when we discuss how to quantify housing surplus or shortage. But first, let’s cover construction.
Mark Fleming - When people track new supply, we usually look at three stages: permits, starts, and completions. Permits are local government approvals that allow construction to begin. Starts mean that a builder has broken ground. Completions mean the home is finished and ready to be occupied.
Odeta Kushi - According to the US Census Bureau's 2024 data, the average time from permit to start was 1.4 months for single-family homes. It's a little longer for multifamily properties at about 2.7 months. That 1.4 months for single-family is down a bit from the 2023 high of 1.5 months, but up from a low of 0.7 months in 2013.
Now from start to completion, the average time for single-family homes was 7.7 months and 16.9 months for multifamily. That single-family timeline is down from last year's 8.6 months but higher than the 2013 average of six months.
Mark Fleming - It’s important to remember that not every permit becomes a start and not every start becomes a timely completion. Financing issues, labor shortages, or material delays can slow things down. I remember during the pandemic not being able to get appliances to finish houses.
Odeta Kushi - Appliances, garage doors, all sorts of issues back then. It’s also helpful to think about this as a supply pipeline. You can have a lot of units under construction and many completions coming to market while builders pull back on permits and breaking ground on new projects. That can mean a short-term oversupply followed by an undersupply.
Mark Fleming - That sounds a lot like the story in multifamily construction right now. The multifamily construction backlog, which we calculate as the ratio of completions to units under construction, is still elevated. That indicates a small backlog at the moment. Last year the headlines were all about elevated multifamily completions, meaning some markets were oversupplied. But permits and starts were trending lower. Once those units are delivered, we could find ourselves in an undersupply scenario if new permits and starts don’t pick up.
Odeta Kushi - That’s a great point. When we look at the latest starts and permits data for multifamily and smooth it with a moving average since it's such a volatile series, we see that permits and starts have been trending sideways, suggesting some stabilization in multifamily groundbreaking. But since we're talking about homebuilding levels, something I always like to hammer home.
Mark Fleming - Okay, nails, hammers.
Odeta Kushi - I see your face.
Mark Fleming - I’m going to one-up you here. MC Hammer? You can’t touch this?
Odeta Kushi - Is that why you were playing that before we started recording? I know the song, but technically it’s from January 1990, not quite the ‘80s. But I'll round up and count it. Really, when we talk about how much we're building compared to historical levels, it's important to look at those numbers relative to the population.
Mark Fleming - Exactly. A million housing starts today isn’t the same as a million in 1980 because our population is much larger now.
Odeta Kushi - Right. We always need to adjust for population growth. Using historical US Census data, we calculated housing starts per thousand households going back to 1920 and found some interesting trends.
Mark Fleming - Starting from the beginning, the average number of housing units built per year rose from 247,000 in 1920 to 785,000 between 1921 and 1928. That was the 1920s building boom, partly driven by a strong economy. But from the late 1920s to mid-1930s, building declined because of the Great Depression. In 1933, housing starts per thousand households fell to just three. That’s a series low but understandable given the economic climate.
Odeta Kushi - By the late 1930s and early 1940s, we were back to building anywhere from 500,000 to 700,000 units a year, about 15 to 20 units per thousand households. Then came the big post-WWII building boom in the 1940s and ‘50s, peaking at 1.9 million units in 1950 or 44 units per thousand households. We’ve never built at that rate since. Going from three per thousand in 1933 to 44 per thousand in 1950 is a huge increase.
Mark Fleming - Exactly. Even the scale—1.9 million units—was massive. That boom was spurred by the GI Bill, which guaranteed home loans with no money down, and the construction of interstate highways, which helped millions of veterans buy homes and move to newly built suburbs. I’m thinking of The Wonder Years.
Odeta Kushi - Maybe a little before my time, but sure. Fast forward to your favorite decade, the ‘80s. In the ‘80s, we were building about 17 units per thousand households on average. There was a dip in 1982 when interest rates hit 18 percent, but even then we built about 13 units per thousand.
Mark Fleming - From the mid-1980s until the mid-2000s, we averaged 15 to 16 units per thousand households. But during the trough of the global financial crisis from 2009 to 2011, we dropped to just five per thousand—almost back to that Great Depression low. As of 2024, we're building only about 10 per thousand households, roughly the same as in 1991. We've been underbuilding since the GFC trough.
Odeta Kushi - There are different ways to measure the extent of underbuilding. Remember we discussed stock and flow measures at the top of the episode? The example we just gave is a flow-stock analysis, with housing starts as the flow and the number of households as the stock. Even when we do a flow-flow analysis, we find we've been underbuilding for over a decade.
Mark Fleming - That flow-flow analysis compares new housing units completed to new demand for housing. Refer back to our first summer school session for a detailed explanation of household formation, which is how we define new demand in that approach.
Odeta Kushi - Right. That analysis measures the housing deficit by comparing new household formation—both rental and owned—to total new housing units completed and added to the stock. This includes single-family, multifamily, and manufactured housing, and we also account for replacing old, obsolete units. We find that month after month in recent years, more households are being formed than homes are being completed. If we pick a point in time when the market was balanced, we can even calculate the size of the deficit. This analysis shows our housing market has faced a shortage since 2006. Before that, we were actually in a surplus.
Mark Fleming - So in the early 2000s—the early "aughts" as I like to say—we were building more homes than household formation rates supported, essentially creating excess supply.
Odeta Kushi - Exactly. Now, one more way to measure the housing deficit or shortage. And remember, each method has its own challenges and assumptions.
Mark Fleming - That’s what you’d expect from two economists.
Odeta Kushi - Indeed. The last method is the vacancy rate approach. Let’s start by defining what that is. Rental and homeowner vacancy rates measure the proportion of rental inventory vacant for rent and homeowner inventory vacant for sale.
Mark Fleming - That’s right. And it’s reasonable to assume that if there’s an overall housing shortage, you wouldn’t see high vacancy rates. So using these rates, we can calculate a housing deficit. But like the last analysis, it requires picking a period when vacancy rates were considered normal, when supply and demand were balanced. For our analysis, we chose 1993 to 2003.
Odeta Kushi - As of the first quarter of 2025, this vacancy rate approach shows a deficit of just under one million units—specifically 943,000. That’s down from the peak of nearly two million units in 2022, but it’s still a significant shortage.
Mark Fleming - Let’s do a quick back-of-the-envelope calculation. In Q1 2025, we were building at a pace of about 1.4 million units annually. If that rate held and there was zero new household formation, no second-home demand, and no homes lost to demolition, we’d get back to normal vacancy rates in less than a year.
Odeta Kushi - But that’s not realistic. Zero new household formation won’t happen, especially given our first summer school discussion about the strong outlook for household formation.
Mark Fleming - I admit I was being a little dramatic with the "zero" scenario. Household formation is well above zero, so that's not a good assumption. Given that, housing will likely remain undersupplied for at least the next few years. But it’s important to note that the supply of housing isn’t the same as the supply of homes for sale. Homes are durable goods. Once built, they can be resold over and over.
Odeta Kushi - That’s a great segue. We developed a metric called inventory turnover. It looks at the sum of new construction supply and existing homes put up for sale. Rather than just counting inventory, it’s better to measure it as a share of the total housing stock, which has grown a lot over the past 30 years. For example, 2.4 million homes for sale in the 1990s isn't the same as 2.4 million today.
Mark Fleming - I see a recurring theme. We keep adjusting for "inflation" in the housing sense.
Odeta Kushi - Precisely. We’ve repeated this idea of adjusting for context a few times now, so it must be important for the exam.
Mark Fleming - Wait—exam? We have exams in summer school?
Odeta Kushi - Sorry, no exam. I got carried away in my professor role. Let me get back to inventory turnover before I panic the class. The pre-pandemic average for this turnover measure was 2.5 percent, meaning 250 out of every 10,000 homes were for sale. Today, it's about 150 in every 10,000. That’s still historically low, though higher than the pandemic-era low of 106 in 10,000.
Mark Fleming - That improvement is showing up in the price data. Annual house price appreciation is now at its slowest rate since 2012, partly because the increase in inventory is easing price pressures.
Odeta Kushi - Very true. But even with more inventory for sale, we’re still in a structural housing shortage. Plus, inventory growth isn’t uniform. It’s increased more in the South, especially Texas and Florida, while the Northeast and Midwest still face tighter conditions. Which begs the question: why can't builders just build more homes where they’re needed?
Mark Fleming - If only it were that simple. Builders face the five Ls: labor, lots, legal issues, lumber, and lending. All of these make it harder and costlier to build.
Odeta Kushi - These challenges aren’t new. They include skilled labor shortages, limited lots, higher financing costs, rising material prices, and regulatory hurdles. All of these have made it tough to deliver more homes.
Mark Fleming - Builders also face higher financing costs, tariff uncertainty, softer demand due to elevated rates, and material costs that remain over 40 percent higher than pre-pandemic. And there’s the chronic labor shortage.
Odeta Kushi - That labor shortage has been an issue for years. Even with gains in construction jobs, we need more workers. Many are aging out and retiring while the industry struggles to attract younger people.
Mark Fleming - On the financing side, NAHB’s survey of acquisition, development, and construction loans shows lending conditions tightened modestly from late 2024 to early 2025, but tightening was less widespread than at any time in the last three years.
Odeta Kushi - A little bit of a glimmer there. And then there are regulatory costs. An NAHB study from 2021 found that regulations at all levels add roughly $94,000, or about 24 percent, to the price of a new single-family home. That was up 11 percent from 2016.
Mark Fleming - And many of those costs get passed on to buyers as higher prices.
Odeta Kushi - Exactly. We talk about the five Ls because more supply is the long-run answer to housing affordability. More homes where people want to live means more choices, which helps moderate price and rent growth.
Mark Fleming - If we break ground on more homes, step-by-step, we'll close the gap between housing demand and supply.
Odeta Kushi - Step-by-step—was that an intentional ‘80s Eddie Rabbit reference?
Mark Fleming - Of course it was. Although not top of my ‘80s playlist. I had to double-check that one.
Odeta Kushi - I’ll never underestimate your ‘80s reference skills again. That wraps up this episode on tracking housing supply. Thank you for joining us for this REconomy Summer School session. If you have an economics question you'd like us to feature in the future, you can email us at economics@firstam.com. And if you can’t wait for the next episode, follow us on X. I’m @OdetaKushi and he's @MFlemingEcon. Until next time.
Thank you for listening, and we hope you enjoyed this episode of The REconomy Podcast™ from First American. We're pleased to offer you even more economic content at firstam.com/economics. This episode is copyright 2025 by First American Financial Corporation. All rights reserved.
