Latest / Inflation Explained with Fexingo: CPI, Prices, and the Cost of Living for Everyday People / Why Shelter Costs Keep CPI Higher Than Headline
Transcript
- Lucas: So the April CPI number hit 332.4 last week, and a lot of the headlines focused on the headline rate cooling. But if you actually look under the hood, shelter costs are still running at 5.2 percent annual growth. That's basically double the overall inflation rate. Luna: And that shelter line item — it's about a third of the CPI basket, right? So when it stays hot, it really anchors the whole index. Lucas: Exactly. Shelter alone accounts for roughly 33 percent of the CPI. So even if used car prices are falling and gas is flat, if shelter stays sticky, the headline number doesn't budge much. And we saw that in April — core CPI was 335.4, up from 334.17 the month before. The trend is still stubborn. Luna: But here's the thing I've been hearing from listeners: they look at Zillow or Apartment List and see rents actually flattening or even dipping in some Sun Belt cities. So why doesn't CPI reflect that? Lucas: That's the key tension. The BLS measures shelter differently than the private indexes. For rent, they survey tenants every six months. For owner's equivalent rent — that's the hypothetical rent homeowners would pay if they rented their own home — they survey a rotating panel. So there's a lag. A six to twelve month lag, typically. Luna: So when Apartment List shows rents down 0.5 percent month-over-month in Austin or Phoenix, CPI might still show those rents rising because it's capturing leases signed six months ago. Lucas: Right. And the magnitude of that lag matters especially now, because we had this huge run-up in rents in 2021 through mid-2023. Those increases are still washing through the CPI. But the real-time data from Zillow's Observed Rent Index shows annual rent growth basically flat — around zero percent nationally as of May. Luna: So the CPI shelter component should eventually cool, but it's going to take time. How much time? Lucas: The Fed's own models suggest that the lagged effect of market rents feeds into CPI shelter with about a nine-month delay. So the softening we've seen in market rents since late last year should start showing up in the official data around the third quarter of this year. But that's not guaranteed. Luna: What could delay it? Lucas: A couple of things. One: multifamily construction completions are peaking right now — we've added over half a million new apartment units in the past twelve months. That's putting downward pressure on rents in places like Atlanta, Nashville, Denver. But if that supply gets absorbed faster than expected, or if single-family home prices stay high enough that more people rent by necessity, then rents could re-accelerate. Luna: And then there's the owner's equivalent rent side. Home prices are still up year-over-year, even if the pace has slowed. That feeds into OER because if homes are more expensive, the implied rent homeowners could charge goes up. Lucas: Exactly. OER is the bigger piece — it's about 24 percent of CPI, compared to 8 percent for actual rent. And OER is even stickier because it's based on a survey of homeowners asking them what they think they could rent their home for. It's a hypothetical, so it moves slowly. Luna: So when we look at the April CPI, the shelter index rose 0.4 percent month-over-month. That's actually a tick down from March's 0.5 percent. But it's still running at an annualized rate of about 5 percent. The Fed wants to see that number closer to 3 percent to feel confident inflation is sustainably heading to 2 percent. Lucas: And that's why the market sold off a bit after the CPI release — the 10-year breakeven inflation rate is still at 2.36 percent, basically flat. The market is saying it's not convinced shelter is going to cool fast enough. And if shelter doesn't cool, the Fed stays on hold. Luna: Which brings us back to the real world. For someone renting in a city where new supply is coming online — say, Austin — they might actually see their rent drop this year. But nationally, the average renter is still paying more than they were a year ago, and the CPI reflects that national average with a lag. Lucas: And that disconnect creates frustration. People feel like the government is gaslighting them — 'inflation is cooling' but their rent went up $50 a month. The truth is, the CPI is telling you what happened, not what's happening now. For what's happening now, you need real-time rental data. Luna: So if you're trying to predict where inflation is headed, watching Apartment List's monthly rent report or Zillow's index is probably more useful than waiting for the next CPI release. Lucas: Absolutely. And that's actually a good note to end this part on. If walking through the economy with us has made something click — maybe given you a better way to think about your own housing costs — and you want to help keep this show going without any ads or sponsors, we do have a listener support page. It's buy me a coffee dot com slash fexingo. No pressure ever, but it helps us keep digging into these numbers every week. Luna: Yeah, and it's a small way to say this kind of analysis matters. We really appreciate the folks who do. Lucas: Alright, so back to the shelter story. Let's talk about one specific market that's a bellwether: Atlanta. Apartment List data shows rents there down about 3 percent year-over-year as of May. That's a big swing from the 15 percent annual increases they saw in 2022. Luna: What happened in Atlanta? It's not like demand collapsed. Lucas: Supply happened. Atlanta added over 30,000 new apartment units in the past two years — one of the highest construction rates in the country. The vacancy rate went from around 4 percent to nearly 9 percent. Landlords started offering concessions — one month free, waived fees — and that effectively lowered net effective rents. Luna: So that's a great example of how local dynamics can diverge from the national CPI. But the BLS data won't capture those concessions fully because it asks about the contracted rent, not what tenants actually pay after incentives. Lucas: Right. And that's one of the critiques of the CPI shelter methodology. It captures list or contract rent, not net effective rent after concessions. So during periods like now, when concessions are widespread, CPI may overstate actual rent growth. Luna: How big could that overstatement be? Lucas: Some economists at the Cleveland Fed estimated that in 2023, the gap between CPI rent and market rent was about 1.5 to 2 percentage points. That's significant. It means the true shelter inflation might already be lower than what's being reported. Luna: So when people say CPI is wrong, they're not entirely wrong. It's just that the official measure is built for consistency over decades, not for speed. Lucas: Exactly. And that's the trade-off. The BLS methodology is rigorous and consistent — it's the same basic approach they've used since the 1980s. That lets us compare inflation over long periods. But it means they sacrifice timeliness. For real-time decision-making, you need private data. Luna: What about the Fed? Do they use the private data when setting policy? Lucas: They do, but informally. They publish what's called the 'Trimmed Mean PCE' and 'Median CPI' which strip out volatile components. But for shelter specifically, Fed staff have their own models that incorporate real-time rent data to forecast where CPI shelter will be in six months. You can see that in the transcripts and speeches — they reference Zillow data sometimes. Luna: So the Fed knows shelter is overstating current inflation. But they still have to wait for the official data to confirm before they act. Lucas: That's the bind. They can't cut rates based on a Zillow index. They need the official CPI and PCE to show shelter cooling. And given the lag, that probably means no rate cuts until late this year at the earliest, unless something else breaks. Luna: Let's zoom out. If shelter inflation does cool to, say, 3 percent by year-end, what does that mean for the overall CPI? Lucas: Quick back of the envelope: shelter is one-third of the index. If it drops from 5 percent to 3 percent annual growth, that alone would reduce headline CPI by about 0.7 percentage points. That could bring headline inflation close to 2.5 percent, assuming nothing else changes. Luna: Which is still above the Fed's target, but a lot closer than where we are now. Lucas: Right. And that's the optimistic scenario. The risk is that home prices stay elevated and OER doesn't fall as much as rent does. Or that the supply wave abates and rents re-accelerate. But for now, the forward-looking indicators are encouraging — if you look at the real-time data. Luna: So to sum up: CPI shelter is lagging, but real-time rents are softening. The Fed is waiting for the official data to confirm. And in the meantime, your personal inflation experience may differ wildly from the national number depending on where you live. Lucas: Exactly. And that's the value of looking beyond the headline. The CPI is a useful tool, but it's not your personal inflation rate. For that, you need to track your own spending — and your own rent.