The news says "inflation has cooled," yet one look at the grocery receipt and you sigh. Are the statistics lying? No. The official price index is right, and your heavy cart is right too. The two numbers simply measure different things. Where the earlier piece CPI Says +2% but Your Cart Feels Heavier tackled the trap of the reference point, this one goes one step deeper: even inside the single word "groceries," fresh food, packaged food, and dining out push prices up at different speeds and for different reasons.

As the maker, handling the answer data for the grocery game, I keep hitting the same wall. It's the same "food aisle," yet some shelves swing every season while others sit frozen for a year and then jump all at once. The structure that sets the price differs shelf by shelf. Understand that structure and the question "why is it so expensive if it supposedly cooled?" gets a lot clearer.

Same "groceries," three clocks running separately

Unpack a cart and you find three bundles whose prices are set in completely different ways: fresh food, packaged food, and dining out. Statistics Korea itself publishes — separately from the headline CPI — a Living-Necessities Price Index (only frequently bought items) and a Fresh Food Index (fish, vegetables, fruit), precisely because these shelves move differently. What items make up each index is laid out in the Statistics Korea definitions (see Sources).

Fresh food — harvest and weather set the price

Cabbage, radish, apples, mackerel — items that go through almost no processing have prices set almost entirely by supply. One heatwave, cold snap, monsoon, or typhoon shakes that week's harvest wholesale, and when shipments from the source fall, the wholesale price spikes within days. Seasonality is huge too: in season it floods in and gets cheap; out of season it's scarce and dear. So fresh food is the most fickle shelf in the cart, swinging sharply up and down within a single month.

The catch is that we buy this shelf every week. The more often you meet a price, the harder a hike imprints. "The zucchini that was 5,000 won last week is 7,000 this week?" — that experience repeats weekly, so the swing of one fresh-food shelf feels like the whole cart got more expensive. You can trace item-by-item wholesale prices by date through the KAMIS agricultural price service (see Sources).

Packaged food — inputs and FX arrive months later

Ramen, snacks, cooking oil, canned goods — packaged food is the opposite. The price barely moves, then jumps one day, and once up it rarely comes back down. The reason is a lag baked into the price. Even when global inputs like wheat, palm oil, and sugar climb, or the exchange rate jumps, companies ride it out for a while on inventory already bought and contracted volumes. Only when the cost burden piles up do they finally raise the shipping price in one move. So a packaged-food hike is often not "today's event" but "an echo of inputs and FX from months ago."

For example, if a snack's input cost rose this spring, the shelf price might only change in summer or fall. And once raised, the price rarely falls back even if inputs drop again (downward price stickiness). The lag and limits of how the exchange rate reaches food prices are covered more deeply in Exchange Rates and Purchasing Power.

Dining out — rent and labor lift the menu

Restaurant prices run on yet another clock. Ingredients make up a smaller share of a dining-out price than you'd think; rent and labor weigh heavily. So even when vegetable prices dip briefly, the menu rarely comes down, because rent and wages rarely fall back once they've risen.

That's also why dining out feels like it lurches "up all at once." A restaurant doesn't renumber a roll of gimbap by 100 won every week. It absorbs the cost burden for a long time, then, when it can't anymore, raises by 500 or 1,000 won in one step. The price jumps like a staircase, so the felt shock is large. For the same total increase, a lunch that suddenly leapt one day stings far more than fresh food that crept up weekly — that's the staircase effect.

Perceived prices center on "what you buy often," so they split from the index

Here's the decisive reason statistics and perception diverge. The official CPI is a single number that weighted-averages hundreds of items by average spending share. Your mental perceived-price, by contrast, is built around what you buy often. Items like fresh food and dining out, met weekly with vivid hikes, swell in mental weight, while a once-a-year drop on appliances or furniture barely registers.

So when one fresh-food shelf swings on the harvest and dining out jumps like a staircase, the headline index can look calm while perceived prices boil. Neither is wrong — the statistic watches "the speed of the average cart," and you watch "the position of the shelf you reach for often." The statistical root of this gap is unpacked through frequency and salience in CPI Says +2% but Your Cart Feels Heavier.

One more thing. When people expect "it'll keep rising," that expectation itself can push prices up. The Bank of Korea separately surveys this inflation expectation, and hikes on frequently bought items stoke that expectation, forming a loop that amplifies perception.

✍️ Operator's note — Honestly, I've often watched a "inflation is tamed" headline and then frozen at the zucchini price at the store. But building the game data, I realized you can't treat food as one lump. For fresh food the culprit is weather, for packaged food it's the FX rate from months ago, for dining out it's rent — and all three clocks tick differently. So the moment you try to explain my cart with one line, "overall +2%," it's bound to miss. Split it by shelf and you can at least see where it's leaking.

A checklist for reading grocery prices with less confusion

• Split it by shelf — figure out whether the rise is fresh food (weather), packaged food (lag), or dining out (rent and labor), and the cause appears.
• Account for season in fresh food — don't mistake an out-of-season price for the year-round level; it has room to fall back in season.
• Packaged food is an "echo" — what rose now is the result of inputs and FX months ago, and once up it rarely comes down.
• Dining out is a staircase — it jumps in one step, so the shock is large but the frequency is low. Separate the size of the hike from how often you notice it.
• The official index is an average — your cart is not the average cart. If the shelves you buy often differ from the average, so will your perception.

Price instinct is built by guessing for yourself

Understanding grocery prices intellectually is one thing; having the instinct to gauge actual prices is another. In PriceGuess's Shopping Price Quiz, guess the prices of items you buy often, and use the Big Mac Index Quiz to compare how the same food differs by country. Curious why the sticker differs from what you actually pay? Continue to Why the Final Price Differs From the Sticker, and for comparing the same product apple-to-apple, read How to Spot the Truly Cheaper Deal — both sharpen the eye that reads a cart.

This article is educational content about how prices form and does not assert price figures for any specific period. For concrete index and item-level prices, check the latest values from official sources such as Statistics Korea and KAMIS.