HACKER Q&A
📣 galdosdi

Why does the media say inflation is up every month?


I'm posting this on this forum because this is an innumeracy issue and you folks are very numerate. If I discuss this anywhere else the politics and emotions will distract, but the numbers are the focus here:

The last three months, core CPI month over month have been 0.0%, 0.1%, and 0.4% respectively (annualized rates of about 0%, 1%, and 5% respectively, or about 2% overall)

Yet for three straight months every time these numbers come out the NY Times and other media plaster a front page story saying "Inflation is stubbornly still at 8.9%" or something, because they are choosing to use a 12 month rolling average, but they all talk about it as if that's just natural, not a choice (and a questionable choice that is at too coarse a level to see what's happening RIGHT NOW)

Certainly there's plenty of nuance to go around -- inflation has been very low for 3 months, but was rather high for months before that, so it's hardly time to celebrate yet. But the indicators from the most recent 3 months are consistently VERY GOOD if you want inflation reduced, and this part, which is also the more news-worthy part, is not getting any attention.

What's up with this? Mainstream media has always been a little innumerate but this is a whole other level. Is it innumeracy in journalism, ideological bias, general fearmongering to sell papers, or what?


  👤 phpisthebest Accepted Answer ✓
For the most part people care, index, and budget based on their income. For most people that changes on an Annual Basis so looking at price flucuations from July 2021 to July 2022 is a more accurate and proper way to look at inflation as it impacts the normal every day person, than looking at inflation variance from Aug 2022 to Sept 2022.

I care that I was buying beef at $3 / lb last year, and now am paying $6/ lb today, not that I was buying beef at $6/lb last month and it is still $6/lb this month... I want my $3/lb beef back


👤 beschizza
The annualized rate is the standard measure and in itself isn't a problem until it's being used to conceal newsworthy changes to the underlying economic trend.

You've pointed on out one thing (Core CPI month over month) in that respect.

More egregious, though, is when media pose the annualized rate AS the month-over-month rate. This happens far too often. The problem is not usually the stories, but the people writing the headlines: innumerate editors who are interested mostly in sensational brevity.


👤 Mikeb85
YoY is the most common measure because many prices and things that happen in the economy are seasonal. Also lots of economic data and prices are only compiled quarterly or yearly so again, YoY makes the most sense.

MoM can show trends but isn't particularly useful otherwise.


👤 bell-cot
If it bleeds, it leads. You don't get clicks/viewers/readers with "Things are getting better in some complex-but-dull way" sorts of stories.

👤 somenameforme
The reason many economic measurements are based on year over year change (not just inflation) is seasonal trends. Numerous factors change, quite dramatically, on an annual basis. That affects not only the factor changing but many others as well.

As but one example, retail business sales skyrocket in November. But that will, in turn, also impact monetary velocity, which will impact inflation, and so on. Seasonal adjustments are possible, but going year over year gives you the most meaningful data.


👤 ravenstine
All you had to do was visit the very first search result with the domain bls.gov.

https://www.bls.gov/cpi/

> In September, the Consumer Price Index for All Urban Consumers increased 0.4 percent, seasonally adjusted, and rose 8.2 percent over the last 12 months, not seasonally adjusted.

That's where the figure comes from. "12 months" is the key aspect.

Either the media is misreporting its context or people are just not understanding what it means.

Pretty much every source I've come across that sites that year-over figure reports it accurately.

https://www.pbs.org/newshour/nation/u-s-inflation-at-8-5-per...

> Consumer prices jumped 8.5% in July compared with a year earlier, the government said Wednesday, down from a 9.1% year-over-year jump in June. On a monthly basis, prices were unchanged from June to July, the smallest such rise more than two years.

https://www.nytimes.com/article/inflation-us-prices.html

> The government reported on Friday that consumer prices climbed 8.6 percent over the year through May, the fastest rate of increase in four decades.

And no, 8.2% inflation over a year is not a good thing. It's not necessarily the worst thing ever, but that's not a minuscule amount of value to lose from your dollar.


👤 sudden_dystopia
They use 12 months because some of the indicators lag by many months. Some data is only generated quarterly or biannually so looking at a smaller timeline wouldn’t really be helpful. The real issue is that CPI is not a poor metric anyway. It was watered down in the 80’s for political purposes to make them look better(on both sides, not one party or the other). Quite frankly, I don’t trust CPI. In my opinion, the best metric for inflation is your eyes vs the rate of change in prices. If you notice several price increases over a span of several weeks, inflation is a problem.

👤 ozb
Note core inflation (excluding food and energy, which are generally more volatile) has been at least 0.6% (annualized >7%) for 5 of the last 6 months https://www.bls.gov/news.release/cpi.nr0.htm

So arguably the narrative is somewhat justified.

But also, I wouldn't be surprised if many in the media are just blindly reacting to the 12-month numbers (8% bad!)


👤 mjthrowaway1
“Consumer prices in the US were up 0.4% month-over-month in September of 2022, the highest reading in three months, and twice the market expectation of 0.2%. Increases in prices of shelter (0.7%), food (0.4%), and medical care services (1%) were the largest of many contributors to the increase.”

Doesn’t seem all that much rosier to have 0.4% MoM with interest rates where they are and GDP declining.


👤 kristjansson
That is the traditional number reported for inflation. If a news source says “inflation was x%” I understand that to mean “CPI is x% higher than 12 months ago”. It’s still a terrible way to present the information, for the reasons you point out.

👤 outside1234
They also don't announce that it is a global problem and actually Germany has even higher levels of inflation.

And the reason for that is that those "everyone is dealing with inflation" stories don't generate clicks.


👤 cykros
Looking to mainstream general media on economic matters is usually going to land you with some skewed portrayals of reality, even in the event they're not resorting to outright lies or inaccurate statements, either because they assume their audience won't understand, because THEY don't understand, or both. I find that at least resorting to something like Bloomberg or Reuters will get a little bit more context around the numbers that isn't completely bizarro, but there is still a pretty high noise:signal ratio there, especially as half of what comes out on the air is someone's projection going forward (think of it as a fairground full of fortune tellers all showing you how shiny their crystal ball is).

I'd also point out that the figures are YoY (Year on Year) rather than monthly changes, but it seems that's been discussed ad nauseam here so elaboration shouldn't be necessary. The one thing I'll point out there is that this is normal across MOST rates unless otherwise stated (your 3 month CD isn't yielding 3% over the next 3 months, it's yielding an annualized rate of 3%, or roughly 0.75% on the principal over its lifetime. This avoids unnecessarily comparing apples to oranges, but can certainly mislead someone who isn't expecting this to be the way things are done.


👤 miked85
I don't believe the numbers regardless of how they are spun. Anecdotally, prices on everything are much higher than +8.9% from a year ago.

👤 dragonwriter
> Yet for three straight months every time these numbers come out the NY Times and other media plaster a front page story saying "Inflation is stubbornly still at 8.9%" or something, because they are choosing to use a 12 month rolling average, but they all talk about it as if that's just natural, not a choice

There are good reasons (month-to-month volatility and seasonality, among others) that the 12-month trailing (not a rolling average) figure is the headline inflation number, but, yes, the media coverage is horrible, with many articles doing some or all of:

(1) Reporting the 12-month trailing figure in the monthly report as monthly inflation,

(2) Reporting the monthly or annual inflation rate (% change in CPI) from the monthly report as the monthly change in inflation.

(3) Failing to provide, or providing incorrect, context as to the import of total vs. core CPI changes when reporting both.

(4) Showing a bias in which of the three above errors are made, and how they are made, in favor of promoting fear.


👤 hulitu
> Ask HN: Why does the media say inflation is up every month?

It is called propaganda. The aim it to make you afraid and then you can be easily manipulated. Inflation is high. But "media" seems particularly interested in this topic in the last few months. Usually combined with other naratives.


👤 kotlin2
Looking at this: https://ycharts.com/indicators/us_core_consumer_price_index_...

Core CPI July 0.31%

Core CPI August 0.57%

Core CPI September 0.58%

Or 4%, 7%, 7% annualized respectively.

That's definitely higher than the Fed's 2% target.


👤 spaceman_2020
Ummm you just mentioned yourself that core CPI went from 0 to 0.1% to 0.4%.

that’s literally the definition of “increasing”


👤 jdasdf
Because inflation has very high seasonality, and so it makes very little sense to compare it with a different season.

It's not "because it's a bigger number" or to "cause shock", it's simply because its a more consistent and useful way of measuring it.


👤 taolegal
Innumeracy? Sure there's tons of innumeracy in the world, but the measures of inflation are so obviously manipulated to indicate low inflation. Really the only accurate way to get a sense of inflation is expansion of the money supply - they can't really lie about that. Everything else is a measure that, in one way or another, incentivizes politicians and economists to lie about their own performance, prowess and power.

Inflation too high before an election? Change the basket of goods. Substitute quality meat for some highly-processed, highly-subsidized puck of corn/soy garbage.

Another thing that even the supposed numerate geniuses among us frequently ignore: inflation compounds!


👤 trynumber9
Well it isn't decreasing. And 0.1% and 0.4% are still up. Combined with a GDP recession we have a bleak picture. For people who aren't overcompensated software engineers it is rather annoying combination.

👤 ponco
As an Australian, with an economy always experiencing very high inflation (8%) I just immediately question what is the composition of the US CPI. Perhaps it’s not reflective or is disproportionately weighted

👤 jjslocum3
This is a timely question, given the front-page headline on this morning's WSJ: "U.S. Core Inflation Hit New Four-Decade High Last Month." Here's an excerpt:

"Prices excluding energy and food rose 6.6%

The increase marked the biggest annual rise in core prices since August 1982, a sign that persistent increases are becoming entrenched in the economy. The overall CPI rose 8.2% in September, down slightly from the previous month’s gain."

I admit I don't know enough to tease out whatever subtleties you're referencing; can you explain what about this is misleading?


👤 trimethylpurine
You said it yourself.

"Mainstream media has always been a little innumerate[...]"

And now mainstream media has been consolidated to just 4 or 5 conglomerate owners who play especially for politics, despite your desire to avoid that topic here. The part that's whole other level is simply the extreme conglomeration from more than 50 news sources which essentially makes it possible to spread mass marketing synergism unchecked and unchallenged.

Do yourself a favor and stop watching it.


👤 giantg2
"Mainstream media has always been a little innumerate but this is a whole other level."

It's not necessarily the numbers.

They don't want to come out saying inflation looks close to normal when people feel otherwise. Not to mention, I think I did see some articles say the .1% was way unexpected and lower. I also think I saw that .4% was higher than expected.

It gets oversensationalized. That happens with other topics in the media too.


👤 breischl
I think you touched on it in your last sentence. It's not actually innumeracy, it's an example of "lying with statistics" to get attention and/or push an agenda (depending on the source).

Tangentially, thanks for bringing this up. I've personally been vaguely aware of all the "inflation is high!!!!!!" articles and hadn't dug in enough to see the reality prior to this post.


👤 purpleblue
No. The Core CPI should have gone down from the steep interest rate hikes. But they didn't and in fact it's still going up. That's huge. 0.4% MoM increase is 5% inflation annualized and that's just for the core. Other parts of the economy are inflating faster and that's a huge problem. The target is 2%. The Fed needs prices to drop and inflation to abate but it's not.

👤 RyanShook
The main reason inflation is hard to measure is that it is not uniform and CPI is a poor representation. Even if the basket of goods used for the CPI is level there are many goods not included in CPI that are still increasing in cost. I agree that the CPI number may be optimistic but I think most people and news organizations recognize that CPI is no longer a good measure of inflation.

👤 gamechangr
I dont know this to be verified, so perhaps those who are interested could look into this.

I have heard the argument that 3 months before every election and/or midterm election ( Nov 8, 2022) inflation is managed to lower numbers.

That could be one reason.

Usually it's used to improve the annual inflation rate. If that's true then you can comment in Dec and Jan if Inflation rates go up again.


👤 cyral
Core CPI excludes food and energy, so the stats you quoted are for the total CPI and are not Core CPI as you claim. Core CPI is at 0.6% this month or over 7% annually, there is nothing low about that and the media isn't really overhyping it. Inflation is still coming in hot despite the rate increases which is not good.

👤 etchalon
To paraphrase Jon Stewart, because the media's worse bias is towards sensationalism and a bit of laziness.

👤 conductr
I was in a ~12 person group conversation this past weekend and most people genuinely thought the continued reporting of the rolling average was ~10% compounding monthly. I hate being a "know it all" type in casual social situations but it was so disconnected from reality that I had to put on my mansplaining hat and give everyone a lesson on how these things work.

Unfortunately, the media is to blame for this and their motives are aligned with the misinformation. They do a horrible job with almost all numerical reporting.


👤 nope96
If anyone is curious, here are the numbers: https://data.bls.gov/timeseries/CUUR0000SA0

Last 4 months have stayed steady-ish at 296


👤 tchalla
Are inflation numbers usually reported the way you suggest?

👤 mitthrowaway2
Does your employer adjust your wage every month?

If not, you probably care about the year-over-year figure.


👤 former_student1
> Mainstream media has always been a little innumerate but this is a whole other level.

Listen dude, it's not the NYTimes's fault your 1 day options expire worthless and your leveraged ETFs went from +X P&L last December to -X P&L today.


👤 rr808
Do you really think most of the US public really care about details? Prices are going up, things are expensive, people click outrage headlines.