Latest / Inflation Explained with Fexingo: CPI, Prices, and the Cost of Living for Everyday People / How Used Car Prices Warp the CPI You Actually See
Transcript
- Lucas: So the S&P 500 just hit 7,483, gold is flirting with $4,200, and silver had its best week in months — up 8 percent. But if you walk onto a used car lot right now, you might feel like none of that matters. Luna: Because the car you're looking at costs roughly the same — or more — than it did two years ago, even though CPI says used car prices have been falling. Lucas: Exactly. The latest CPI print — May 2026 — shows the used cars and trucks index down about 4 percent year-over-year. But that doesn't match what real buyers are seeing. Dealer auction data from Manheim shows wholesale prices actually ticked up in June. So there's this gap. Luna: And that gap is exactly why a lot of listeners write in saying 'CPI doesn't reflect my life.' Today we're going to drill into one specific reason for that gap: the hedonic quality adjustment. Lucas: If walking through the economy with us has made something click — maybe around how inflation is measured, or why your grocery bill doesn't match the headlines — a handful of listeners chip in monthly at buy me a coffee dot com slash fexingo. That's literally what keeps this show ad-free and lets us spend time on nerdy corners like hedonic adjustments. Luna: Yeah, it's a small group that makes it possible for us to do episodes like this one. And we're grateful for every single one. Lucas: So back to the hedonic adjustment. The Bureau of Labor Statistics — the folks who compile CPI — have a problem: cars change every year. A 2025 Camry has slightly different features than a 2024 model. The BLS tries to strip out the value of those improvements so that they're measuring pure price change, not the cost of getting a better product. Luna: Makes sense in theory. But in practice, it can make it look like prices fell when they didn't. Lucas: Right. Let me give you a concrete example. Imagine you're looking at a 2022 Toyota Camry LE. That car had a starting MSRP around $26,000 new. In 2024, a similar used one might have sold for $28,000 because of the supply crunch. Now fast forward to July 2026. A 2022 Camry with 40,000 miles is still listed at $27,500 on CarGurus near me. Luna: So the cash price is basically flat from two years ago, maybe down a tiny bit. Lucas: But CPI's used car index might show that car's price fell 4 percent. How? Because the BLS is comparing a 2022 Camry to a hypothetical 'base' Camry that has been adjusted for quality. They might say the 2022 model lacks a feature that became standard in later years—like a bigger touchscreen or better lane-keeping assist—so they subtract a few hundred dollars of 'quality' from the 2022 car's price. Luna: Which makes the price appear to drop, even though the actual transaction didn't change. Lucas: Exactly. And this isn't just a theoretical quirk. A 2019 paper by the BLS itself found that hedonic adjustments for used cars can account for up to half a percentage point of downward bias in the index per year. That's huge when you're talking about a component that makes up about 4 percent of the CPI basket. Luna: So the CPI says used car prices are falling, but in reality they're just not rising as fast as they would have without the adjustment. Lucas: Or they're rising, but the adjustment masks it. Let me pull in a real number from the data we have. The CPI for all urban consumers hit 334.0 in May. Core CPI—excluding food and energy—was 336.1. Those are both up about 0.5 percent from April. But used cars? The index actually dropped 0.7 percent month-over-month. Meanwhile, gold is up 4.1 percent in the last five days alone. Luna: So investors are piling into hard assets because they don't trust the inflation numbers? Lucas: Partly. But also because the labor market is softening—the June jobs report showed only 57,000 new payrolls, and the unemployment rate ticked up to 4.2 percent. That makes people nervous about growth, so they seek safe havens. But it also means the Fed might cut rates sooner, which is good for gold. Luna: And if the Fed cuts rates, that could actually boost demand for used cars—cheaper financing—which would push prices up further, widening the gap between CPI and reality. Lucas: That's the loop. The Fed looks at CPI and sees inflation moderating. But real-world prices for big-ticket items like cars are sticky. The hedonic adjustment makes CPI look lower than what consumers experience. So the Fed might ease too soon, which reignites inflation in other areas. Luna: And that's exactly the kind of thing our listeners should understand when they hear 'inflation is under control.' Lucas: Yeah. So let's get into the weeds on how the hedonic adjustment actually works for cars. The BLS collects data on specific models with specific trim levels and mileage. They track a sample of vehicles over time. When a new model year comes out, they compare it to the old model year and estimate how much of the price difference is due to new features. Luna: How do they estimate that? Do they just guess? Lucas: No, they use regression models. Basically, they look at a bunch of car sales and figure out how much each feature—like a sunroof or leather seats—adds to the price. Then they apply those coefficients to the new model to strip out the value of any added features. So if the 2025 Camry gets a $500 upgraded infotainment system, they'll subtract $500 from its price before comparing it to the 2024 model. Luna: That sounds reasonable, but it assumes the market values those features exactly the way the model predicts. Lucas: Right. And that's where it breaks down. In a supply-constrained market—like we had in 2021-2023—buyers were paying over MSRP for base models with no options. The hedonic model might say a feature is worth $500, but in reality, buyers would pay $2,000 just to get any car. The model doesn't capture that scarcity premium. Luna: So the adjustment makes the price look lower than it actually is, because it's subtracting 'quality' that the market doesn't even price in at that moment. Lucas: Bingo. And there's another subtlety: depreciation. The BLS assumes that a car loses value over time due to age and wear. But in recent years, some used cars have appreciated. If the model assumes a 15 percent depreciation per year, but the actual depreciation is 5 percent, the index will show a price drop that doesn't exist. Luna: So the hedonic adjustment is working fine in normal times, but it's terrible at handling the kind of market dislocations we've seen. Lucas: Exactly. And the BLS knows this. They've been tweaking the methodology. In 2024, they started using a new 'quality adjustment' process for used cars that incorporates more data from online listings. But the fundamental issue remains: you're trying to measure something that doesn't have a clean benchmark. Luna: What about new cars? Does CPI handle them the same way? Lucas: New cars also get hedonic adjustments. The new car index was up about 0.3 percent in May. But again, if you look at transaction prices—what people actually pay—they're often higher because of dealer markups and low inventory. The CPI new car index has been rising modestly, but not as much as the real out-the-door price. Luna: So the net effect is that both new and used car inflation are understated in CPI. Lucas: That's the takeaway. And for a listener shopping for a car right now, what does that mean practically? It means don't assume that because CPI says car prices are falling, you'll get a deal. You might get a slightly better price than a year ago, but compared to 2020, you're still paying 30 percent more. Luna: And if you're trading in a car, the same logic applies in reverse. The trade-in value might be higher than CPI suggests because the hedonic adjustment is also suppressing the value of older cars. Lucas: Right. So for listeners looking to buy or sell, the best guide is not CPI but auction data or sites like Manheim's used vehicle index. That's closer to real-time market pricing. Luna: And that's the value of understanding these measurement quirks—you stop relying on a single number and start looking at the underlying data. Lucas: Yeah. And speaking of underlying data, we should note that the 10-year breakeven inflation rate is currently at 2.23 percent. That's the bond market's expectation of average inflation over the next decade. It's been flat, which suggests investors aren't panicking about inflation spiking again. But if the hedonic adjustment is masking real price pressures, then the breakeven might actually be too low. Luna: So the bond market could be getting a distorted signal from CPI? Lucas: Potentially. And that's why the Fed looks at multiple indicators—PCE, core PCE, the trimmed mean from the Dallas Fed. They know CPI isn't perfect. But the average person hears 'inflation is at 3.4 percent' and thinks that's what they're experiencing. We're trying to help bridge that gap. Luna: And maybe the next time you see a headline about used car prices falling, you'll think twice. Lucas: Exactly. Alright, that's it for this episode. We'll be back tomorrow with another angle on what inflation data actually means for your wallet.