Hacker News Reader

Comment by PragmaticPulp | Hacker News Reader

From the article:

> In the U.S., prices rose close to 3% in the year ending Jan. 2, according to NielsenIQ, roughly double the overall rate of inflation.

The study found that food prices are rising faster than inflation.

3% is high, and 1.5% over inflation is also concerning, but I think a lot of people are jumping to conclusions based on the headline. Those inflation numbers would go unnoticed by most during normal times. We’re only seeing these headlines because inflation is a popular media topic right now.

View on HN

Disagree that this is alarmism. The question is whether the way inflation is being measured matches the way we use it to make decisions. Graphs like AEI's Chart of the Century [1] have been highlighting for years that inflation numbers are inappropriately averaging decreasing costs for discretionary purchases, while costs for essentials like healthcare and education grow out of control. When food also starts outpacing the average (historically it has paralleled wages in the USA), you really have to ask what inflation is measuring.

Low inflation is essential to justifying the monetary easing policies that have been sustaining the stock market through the pandemic... so it is important for everyone to keep an eye on how dialog changes around these numbers.

[1] https://www.aei.org/carpe-diem/chart-of-the-day-or-century-3...

The food inflation number was isolated in the article to be around 3%.

Much of that increase can be traced back to increased shipping costs during COVID lockdown.

Monetary policy isn’t the only or even the primary driver of inflation. In this case, the inflation comes from increases demand on a portion of the food supply chain (shipping services).

The alarmism in the comment section largely comes from people misreading the headline. The article is about soaring food costs in developing nations, but the comment section is treating the headline as if it’s about US food prices.

The article also discussed the fact that US prices are temporarily buffered by the supply chain but that big providers like GM are discussing price increases do to sustained cost increases. This is alongside existing practices like shrinkflation that make it harder for a CPI to accurately benchmark costs. It makes sense that articles and readers are looking for leading indicators.

> Monetary policy isn’t the only or even the primary driver of inflation.

After two degrees in economics, my perception is that monetary policy is definitely considered a primary driver of inflation, probably THE primary driver. But this is probably irrelevant - see below.

> In this case, the inflation comes from increases demand on a portion of the food supply chain (shipping services).

Yes, for example: "in North America... a shortage of both shipping containers and truck drivers".

> The alarmism in the comment section largely comes from people misreading the headline.

To me, it feels alarmist for Bloomberg to consistently say "inflation" instead of "price increases" throughout the article. A temporary problem with supply/demand is not the same as inflation. Food prices will almost certainly fall back to their normal trajectory within a year or two, once supply/demand normalizes, whereas if there was truly inflation that would almost certainly NOT occur.

Education and healthcare have pathological cost growth in large part because they aren’t subject to ordinary market pressures (the consumers are either over a barrel and can’t say no or aren’t paying in the first place). We need to fix them but it would be a mistake to let those pathologies drive monetary policy.

I don't have many proactive opinions about monetary policy, but it's worth noting that health care and education together make up 25% of US dollar-valued GDP and significantly more than that of GDP growth. If their pathologies enable them to successfully suck up spare dollars, that is a meaningful hole in the bucket for efforts meant to impact other behaviors.

> Graphs like AEI's Chart of the Century [1] have been highlighting for years that inflation numbers are inappropriately averaging decreasing costs for discretionary purchases, while costs for essentials like healthcare and education grow out of control.

To be fair, three out of the seven categories that are under inflation can also be considered "essential". Imagine the outrage if cars (or more generally, transport) and clothing increased 2x over inflation, or if cellphone/internet access became less and less affordable (remember all the "internet should be an essential service" headlines at the start of the pandemic?).

yeah, i don't love my discretionary/essential breakdown there. it's more about the combo of essential + non-exchangeability -- you need a car but there are a lot of options, but when you need a hospital there's usually one best place. (and education is non-exchangeable due to competitiveness).

the other big correlate is just goods vs services (even textbooks are service-like due to editions). I am not an economist but I'm really curious whether rapid depreciation caused by planned obsolescence of e.g. consumer electronics is causing a poor measurement of price trends in that area. Today's iPhone will cost much less next year, but the cost of maintaining a cutting-edge smart phone keeps going up.

> When food also starts outpacing the average (historically it has paralleled wages in the USA), you really have to ask what inflation is measuring.

inflation is measuring the value of a currency. CPI is a way to infer inflation from the nominal price of consumer goods; but it is not the definition of inflation. put another way, if an apple costs twice as much as it did last year, that fact is not enough to say whether the currency inflated.

"the value of a currency" isn't a physical reality; it is only observed through price fluctuations. so, defining it this way is just kicking the can. The fact that you can measure very different values by focusing on Forex, labor costs, or various CPI baskets is exactly the concern with gaining too much comfort from any single-number summary in the face of price changes that are impacting daily life for millions.

I would consider the value of currency to be as real as the value of a software license, even though neither is a physical item per se. I guess that's somewhat philosophical.

my real point here is that nominal price increases do not necessarily imply inflation. if the price of one type of good is increasing much faster than the others, a better explanation is that the real value of those specific goods is increasing. does this matter to ordinary people buying food? not really. their main concern here is the nominal price of food compared to their nominal earnings.

Do we have a way of seeing periodic media obsessions? Seems like a good AI challenge, although would require a good corpus.

I saw another analysis suggesting a 75% drop in media coverage of Biden vs. Trump, at this same point in each’s presidency. That’s a useful statistic that I have a hard time finding in a comprehensive way.

>> Do we have a way of seeing periodic media obsessions?

We have dozens. Countless companies have tools to show you how things are happening on social media, and all the big new outlets mirror their content to social media. At the most simple level, basic word counts can track how issues appear and disappear. When you hear about a 75% drop in coverage of "Biden v. Trump" that is almost certainly taken from the number of times those two names are included in news reports. Counting words or highlighting the use of new words does not require AI.

Topic modeling is a better technique than simple word counts. Topic modeling can use ML.

Reply on news.ycombinator.com