"Nationally, Housing Is Unaffordable": A Conversation on Housing Affordability
Jessica Dill: Welcome to another Federal Reserve Bank of Atlanta podcast. I'm Jessica Dill, director of the Atlanta Fed's Center for Housing and Policy. I'm here today talking with Domonic Purviance, a senior financial specialist who also has expertise in residential real estate. Hi, Domonic—thanks for joining me today.
Domonic Purviance: Hey, Jessica—glad to be here.
Domonic Purviance and Jessica Dill of the Atlanta Fed
Dill: I have you here today to talk about your recently launched Atlanta Fed Home Ownership Affordability Monitor, or HOAM, tool. When did you introduce the tool, and why did you create it?
Purviance: Well, the HOAM tool was introduced at the beginning of this year, and it's something that we had been using internally. I work in banking supervision here at the Federal Reserve Bank of Atlanta. We track home ownership affordability normally as a leading indicator of the health of the housing market. Typically, what happens is housing becomes less affordable, and then you have some sort of softening of credit. And so we pay attention to housing affordability, because it lets us know the trajectory of where we are and the overall health of the market.
We decided to make it public for a couple of reasons. First, we want to assist the public in assessing housing market conditions from the perspective that we have internally at the Federal Reserve. But we saw that some of the tools that existed currently on the market did not include some other variables that we thought were very important in order to track housing affordability—most notably, being able to look at affordability at lower levels of geography, including down to the county level. And so we believe that introducing this tool and making it public will assist people that are examining the housing market and allow them to better understand what affordability conditions are like, particularly at lower level geographies.
Dill: What does the HOAM tool measure, and—importantly—what doesn't it measure?
Purviance: Affordability is sort of a tricky thing to measure, surprisingly. You would think it'd be a little bit easy, but there are several things that go into looking at affordability. First of all, our measure looks at what the median-income household can afford given the median-priced house and the current interest rate. In addition to looking at the principal and interest payment, given the current interest rate, we also include the PMI [private mortgage insurance]. We also look at taxes and insurance. Those measures collectively encompass what a total housing cost would be for a household per month. We also estimate a down payment of about 10 percent when we calculate the principal with interest payment. When you add that together, the total cost—the PMI, the principal with interest payment, taxes and insurance—the affordability for us is 30 percent of income. So if your total housing costs per year exceed 30 percent of your income, then that particular market or geography is considered unaffordable. If your total cost is below 30 percent of income, it's considered affordable. And so that's what it measures. It measures affordability based on the share of income that your total housing costs cover per year.
What affordability doesn't measure is, number one, how expensive a particular market is. For example, there could be a market that is more expensive but actually is not necessarily unaffordable, because remember, affordability takes into account what the median-income household can afford. So you could have a market, for example, like San Francisco, that's one of the most expensive markets in the country, but because incomes are higher in San Francisco it still may not be as unaffordable as say, a market like Los Angeles—where incomes are a little bit lower—even though housing in Los Angeles is less expensive than San Francisco. And so again, affordability measures the share of income that goes to housing, but it doesn't necessarily measure how expensive a market is. It also doesn't measure available inventory, so you could have a situation where a market is relatively affordable, but it doesn't mean that there is inventory available at a certain price point that a household can afford.
And lastly, it doesn't necessarily measure affordability for all households. If you make the median income in a particular region, you may be able to afford to purchase a house, but if you make below the median income, then housing may not be as affordable for your household. And so those are the main things that our affordability tool does not measure.
Dill: And just to be clear: when you say it doesn't measure affordability for all households, this particular tool doesn't measure affordability for renter households, right?
Purviance: Right—we're just talking about home ownership. We do have some other tools that look at renters at the Federal Reserve Bank of Atlanta, but this particular tool is just looking at home ownership affordability.
Dill: Can you tell us a little bit more about the intended audience?
Purviance: The primary audience, as we see it, would include public and private entities that are involved in evaluating housing-related issues. So it could be anything from a researcher at a university or a local government that's trying to evaluate market conditions to, obviously, entities like the Federal Reserve or banks or home builders, that are trying to evaluate housing market conditions and trying to make decisions about where to lend or where to build or develop. It does have some utility for the average home buyer that's out there that is looking maybe to move in a certain market or to make a decision on where to buy, and they are looking just generally where housing is more affordable given their current income. But generally, our intended audience is more entities that are evaluating overall market conditions.
Dill: And just to circle back on that last point—it won't necessarily help someone who's looking to buy a house and tell them whether they can afford a house—right? This is more for folks that are maybe considering a metropolitan area and trying to see which part of the metropolitan area is more affordable than others—generally speaking, not for their specific circumstance.
Purviance: Right. There is some nuance, but this is more of a "30,000-foot-level" tool. Now if you make exactly the median income, it may tell you where you could afford to buy, but it's highly unlikely that the average home buyer looking at this tool is going to have. They're going to be above the median or below the median, and it gives you a general take on the affordability in a market. And it could very well be that a market may be unaffordable based on the median income, and you may make below the median income. It doesn't mean that there aren't homes available in that market for you to purchase. Again, whenever you're looking at the median or some measure of the central tendency, it doesn't tell you everything, but it just gives you what the center of the market looks like, and then the "30,000 foot" view. And then you drill down to help you make decisions if you're a home buyer or a builder-developer or some of the other entities we mentioned.
Dill: Thanks, that's helpful. How would you say the HOAM tool is different from other affordability tools that are out there? What sets it apart, and why should someone come and use the Atlanta Fed HOAM tool versus another organization's tool?
Purviance: Well, first I'll say that there are a lot of tools out there that measure housing affordability, and they all have some measure of utility and benefit. We didn't develop the tool because we thought that there aren't already good measures of affordability out there, but our tool has some nuances that I think make it distinctive and that perhaps provide some things that some other tools don't provide. So the first important distinction, I think, is that we believe our measure of affordability takes into account some cost of home ownership that some other tools may not include. As I mentioned before, we not only look at principal and interest payments, we also look at taxes and insurance as well as PMI. And for the average household, when they're thinking about affordability they're including taxes and insurance and other things beyond just the principal and interest payment.
The other thing that makes our tool distinctive from other tools is we believe we make some more nuanced assumptions, particularly around down payment requirements. There are a lot of tools that use a 20 percent assumption on down payment, but when we were doing the analysis we saw that in the universe of home buyers the average down payment is closer to 10 percent, and so we use a 10 percent down payment because we think it's a little bit more realistic to the average home buyer. The other thing that makes our tool distinctive is that we use household income versus family income. This is a little in the weeds, but there are some tools that look at the median family income to measure affordability. Now, "families" are defined as two or more related people that live in a household, so you tend to get more couples—dual incomes—and so the family income tends to be a little bit higher.
However, when we were doing the analysis, we found that around 27 percent —somewhere between 25 and 30 percent —of home buyers are actually single-person households. And so if you're just looking at family incomes, at least in our estimation you will be excluding about a third of the market. And so we use household income, which includes families but also includes single person households that are a good portion of the home buyers. Lastly, and probably the most important distinction in our tool, is we go down to lower-level geography, so we're looking at counties, versus just metro areas or the nation as a whole. If you're doing an evaluation of a market, the overall market may be affordable but you may have some counties within the market that are maybe unaffordable. This was done for us internally—when we're looking at risk, we have to look at risk at even below the county level—down to the ZIP code level, or even the census tract level. So that gives you a little bit more nuanced picture of what's going on. And of course we update our tool monthly, and there are some other tools that have less frequency. We believe that all of these things make our tool a little bit more distinctive.
Dill: Thanks—I think that's really helpful in understanding the landscape of what's out there, and then how our tool fits into that landscape. I wanted to move on from the actual tool to what the tool is telling us. As you mentioned in your earlier remarks, you said that we released the tool earlier this year—and for our audience, we actually released it in the middle of March 2020.
Purviance: Perfect timing, huh?
Dill: Perfect timing with when the coronavirus hit. And so we put out a lot of press releases, but we're not sure it really rose to the top of everyone's radar of what's happening because everyone was really concerned about what's happening with the coronavirus, and what impact is that going to have on households' income and on their housing situation. And so since coronavirus has hit, we've been tracking home ownership affordability through our HOAM tool. What has the trend been in homeownership affordability? Has there been any movement in home prices, interest rates, or household income that's been surprising to you?
Purviance: Yes, actually. As many of our listeners might be aware, interest rates have dropped considerably since the beginning of the pandemic, and any time interest rates drop, that creates increased affordability for home buyers. However, there are several different variables that are all moving at the same time, so as interest rates declined, home prices have actually increased. The reason why home prices have increased is related to inventory levels. So just to explain, there was early on—at the end of March and beginning of April—a contraction in demand. There were a lot of people that were thinking about buying that delayed purchasing. If you look at pending home sales, they declined sharply at the end of March, beginning of April, but at the same time, inventory declined. So there are a lot of people who were thinking about selling, but they moved their home off the market. As demand has recovered, around the beginning of May and going into June, there just hasn't been enough inventory. And so the lack of inventory has created upward pressure on home prices. If you look nationally at the median home price, the U.S. is somewhere around…if you look at the three-month moving average, in May it was somewhere around $290,000, the median home price. That's higher than it's ever been, and so even though interest rates are going down, home prices are going up, and so the net effect has been actually a decline in housing affordability.
Our latest measure in May, that we just actually finished calculating yesterday, shows that nationally, housing is unaffordable. And it is probably a more conservative estimate, because we are in the process of reprojecting our income assumptions. Whenever there's a recession, you can expect households to experience a contraction in incomes. And so those two factors—home prices going up, and income declining—have been more than enough to offset the decline in interest rates and have made home affordability decline overall. And just one other factor to mention: credit has also tightened, and so there have been higher down payment requirements from many lenders and also higher credit requirements. And so the people who are buying now are people that have maintained their income through the pandemic and also are in a good credit position to go buy. If you look to home builders, it has had a positive impact on demand overall. Demand year over year is much higher in May and June than it was in 2019, but it's difficult to say how well that will be sustained moving forward now that we see conditions with the coronavirus turning a little bit for the worst. So right now, in terms of housing affordability, so far the net result has been, through the pandemic, a decline in overall housing affordability.
Dill: So there's been a decline on a national level, but have you observed any divergence regionally in affordability trends across the country since COVID hit? With our tool, we can go in and look at different geographies and drill down, and so I'm curious what you're seeing on that front.
Purviance: Yes, I think overall around 80 percent of the markets that we include in our survey are still affordable, just looking at our measure—although you see the affordability declining there still, in terms of the percent of income required to afford a house is still a little bit below 30 percent in most markets—and about 20 percent of markets were considered unaffordable. And it's not a surprise where those markets are: most of your coastal markets in California, in the Northeast, as well as in South Florida, continue to be unaffordable. In our district in particular, in the Southeast, the only market that stands out in terms of a decline in affordability, surprisingly, is Augusta, Georgia, where affordability declined about 3 percent—2.8 percent—in our latest index. And that was primarily because of, as we mentioned before, an increase in home pricing.
Dill: Since COVID hit, there's been a lot of talk comparing and contrasting this downturn with the Great Recession. As we think about affordability, does the HOAM tool provide a historical reference point? And what can we glean from our tool in terms of how this downturn compares to the Great Recession, specifically with regard to affordability?
Purviance: Our tool goes back to 2006, so it would include—at least nationally, it goes back to 2006—so we can get a look at what was happening prior to and during the Great Recession. The big difference is that the Great Recession was a housing market–related recession, at least in part, where with this recession the downward trend is unrelated to housing. So it's quite different, and if you just look at affordability, what happened prior to the recession in 2008, if you go back to 2006 our tool shows that housing affordability had contracted considerably. And so we had several years prior to 2008 where housing affordability was below the affordability threshold, meaning more households were spending more than 30 percent of their income on housing per year at the same time credit loosened considerably. The last crisis was a subprime mortgage crisis in particular, and so as credit softened eventually we had a rise in delinquencies and defaults, and that really precipitated the downward spiral in housing. It led to an increase in inventory, and put downward pressure on home prices. What we're seeing now is quite different from that. From the last crisis, we have not seen an increase in subprime mortgage originations. So on that side of it, we haven't really seen the deterioration of credit. However, where we have seen a weakness is on the debt-to-income side, and so we've been trying to make the point for the last year or so that even though we didn't see an increase in subprime mortgages, we did see a deterioration in debt-to-income ratio. So more households were spending a greater portion of their income on housing, and that was a direct result of a decline in affordability in many markets.
As I mentioned before, in the Great Recession we saw an increase in inventory. This time we're not seeing an increase in inventory at all. Prior to the recession affordability had declined, and then during the recession—because of the increase of foreclosure inventory—we saw a pretty big spike in inventory as well as downward pressure on price. We're not seeing any of that today. Right now, prices are actually going up. We're seeing less inventory on the market. However, the one thing that we are paying attention to that might be a risk moving forward is—and we really didn't have a whole lot of this in the last recession—is an increase in forbearances, in government support that forestalls defaults. So the average homeowner that's hit a forbearance, based on the CARES Act you have up to a year before you have to start making payments, theoretically. So far, it seems like the forbearances and government stimulus has helped to forestall defaults, unlike what we saw the last time. But as we continue, that may not be sustainable.
Dill: Thanks, Domonic. This has been an enlightening conversation about the HOAM tool, recent affordability trends, and COVID-19-related impacts on the housing market. As we close, I'd like to put in a plug for the Atlanta Fed's Center for Housing and Policy: In addition to offering a suite of data tools, which includes the HOAM tool, we've also published several Real Estate Research blog posts that explore current market conditions, mortgage forbearance uptake, and vulnerable renter households. Check back often for more HOAM tool updates, which come out on the 15th of each month, and for more Real Estate Research blog posts covering COVID-19 impacts on the housing market. Thanks for tuning in to another Atlanta Fed podcast—and thank you, Domonic.
Purviance: Thank you.