Ed Dolan's Econ Blog

Inflation is Quiet, So Why are People Still Feeling its Pain?

This week’s report from the Bureau of Labor Statistics shows no change in the seasonally adjusted U.S. consumer price index for April. Real average hourly earnings were also unchanged. On the face of it, those numbers should take inflation off the list of things people have to worry about, but they don’t. Instead, every time I post numbers that show inflation is low, I get comments like these:

“Inflation is creeping in. Businesses are holding back. People are up against the wall.”

“Those who live in ivory towers and pontificate about the ‘real’ economy while being totally removed from it should really experience how the 99 percent live.”

“If you believe the BS stats, you are in the minority. Consumers KNOW differently. Core CPI, etc. What a pile of cowdung!”

It is tempting to dismiss comments like these as simply ill informed, but that would be a mistake. Instead, we should take them seriously as an indication that when economists and consumers think about inflation, they ask different questions and get different answers.

The differences all stem from the fact that consumers look at what affects their pocketbooks–actual changes in prices as they happen. Economists, who instead are looking for what is relevant for policy, adjust the raw data to eliminate statistical noise that obscures underlying trends.

Seasonal adjustments are one example. When the BLS reports that the seasonally adjusted CPI for April did not change, it is giving an answer to the question, “Did prices change more or less last month than they usually do in April?” Only changes that differ from the April norm are relevant for policy. Policy operates on a time frame much longer than one month. There would be no point in adjusting interest rates or government expenditures just to offset price changes that happen each April as a matter of course.

In contrast, when consumers visit the supermarket or the gas station, they see unadjusted numbers. Just yesterday, I met a neighbor at the gas station who commented, “Wow! Look how much gas has gone up!” I was momentarily taken aback to hear this. Less than an hour earlier I had looked at the latest BLS report and noticed that it showed a 2.6 percent decrease in motor fuel prices for April. But, oops, that was the seasonally adjusted number. The unadjusted price of motor fuel went up by 1.8 percent in April. I realized that I was so used to looking at the seasonally adjusted data from the BLS that I had not even bothered to look at the numbers on the gas pump right in front of me. In fact, the CPI as a whole, not seasonally adjusted, rose at a 3.7 percent annual rate in April even though the seasonally adjusted change was zero.

It would be wrong, though, to think that consumers look exclusively at unfiltered data while economists apply filters. In fact, both filter the incoming data, but in different ways.

In their efforts to filter out statistical noise,  economists  go beyond simple seasonal adjustment and pay special attention to  inflation measures like the core CPI, which removes prices for food and energy from the index. Food and energy prices move around a lot from month to month, so they obscure the policy-relevant trends economists are looking for.

Instead, consumers tend to ignore prices that do not change. As Daniel Kahneman notes in his book Thinking Fast and Slow, the tendency for our brains to focus on things that change or move has deep evolutionary roots. The eagle sees the running rabbit; the rabbit sees the eagle’s moving shadow and freezes, hoping to become invisible. Because they change often, food and energy prices are what consumers notice most. At the same time, they pay less attention to prices like those for manufactured goods, which have less month-to-month volatility. Not long ago, I wrote a post that drew attention to how little inflation there has been for the items found in a typical Sears Catalog—things like clothing and appliances. The response was spirited, to say the least. One comment after another insisted that the true indicators of inflation were milk and gasoline, not shirts and TVs. Yet, according to the BLS, “other commodities,” as a group, make up a larger share of consumer budgets than either food or energy.

There is another, even more important difference in the weights and filters that consumers and economists apply to changing prices. Economists  treat price changes equally, regardless of their direction. As an example, consider another measure of underlying inflation, the 16 percent trimmed mean CPI published by the Cleveland Fed. That measure symmetrically removes the 8 percent of prices that increase most and the 8 percent that increase least each month, regardless of what they are. Since food and energy are not always the most active prices, some economists consider the trimmed mean CPI to remove statistical noise more effectively than the core CPI.

Consumers do not treat price increases and decreases symmetrically. Instead, they give more weight to those that increase than to those that decrease. Economists call this loss aversion. Using the previous month’s prices as reference point, consumers perceive any price increase as a loss and any price decrease as a gain. For changes of equal dollar value, the pain of an increase outweighs the pleasure of a decrease. For example, in applying loss aversion to the April data, we would find that people would feel the pain of increasing prices for fruit and vegetables more than the pleasure of falling prices for dairy products, and the pain of rising motor fuel prices more than the pleasure of the decreasing price of natural gas.

When we put all of these factors together, economists and consumers end up with very different views of inflation. The chart gives a visual representation of their different viewpoints using three versions of the CPI, all showing monthly values as annual rates. For economists, we show the seasonally adjusted core CPI (red) and trimmed mean CPI (blue). For consumers, we show a “perceived CPI ” (green).

To construct the perceived CPI, I began with the all items CPI, not seasonally adjusted, and modified it to simulate the effect of loss aversion. The importance of loss aversion depends on the degree of volatility of prices. As a measure of volatility, I used the mean absolute value of the difference between the monthly trimmed mean inflation rate and the all-items CPI without seasonal adjustment, which was .38 percentage points over the period in the chart. To reflect that fact that losses from price increases tend to be weighted about twice as heavily as gains from price decreases, I divided the volatility measure by two to give a 0.19 percentage point upward bias due to loss aversion.

The green line is certainly a lot more alarming than the red or blue one, but it is still not quite the whole story. One final consideration adds to the difference between the way consumers and economists perceive inflation. As economists see it, inflation causes pain only to the extent that it affects real purchasing power. If inflation raises nominal earnings at the same rate as consumer prices, it should cause no pain at all. That would have been the case in April, when real earnings did not change. Furthermore, as Tim Duy documented in a recent post, the April situation was typical, although there are exceptions in some periods and for some group of workers.

Consumers see things differently. They see increases in nominal earnings not as an effect of inflation, but as deserved rewards for their own skills and efforts. They then perceive any erosion of the purchasing power due to inflation as an undeserved loss. It follows that even if nominal earnings and the CPI rose by the same amount, and even if all prices changed at the same rate, inflation would cause pain.

The bottom line: The inflation measures that economists use may be useful in formulating policy advice, but they do not even approximately measure the psychological pain of inflation. At any rate, that is how I explain the constant stream of comments like “Your calculations seem like a magician’s trick to me,” “political manipulation,” “Gumint lyin’ to us again,” “you just don’t see what’s really going on,” and all the rest.


16 Responses to “Inflation is Quiet, So Why are People Still Feeling its Pain?”

Bryan WillmanMay 18th, 2012 at 2:34 am

remember too that contracts and commitments are normally nominal and not net of inflation, and not indexed to each other very well. so while my rent may only adjust upward once a year, fuel to drive to work may change every day, and my wages may only change every couple of years.
even if these all average out to zero net real change over say three years, i can still be nominally insolvent this month.

llisa2u2May 18th, 2012 at 10:30 am

Well, a sausage egg biscuit at McDonalds is now 2.95USD. and 2 apple pies have now dropped off the $1menu, and are now how much? I think that says it all that real inflation that the day-to-day minimal consumption level consumer is getting hit with is uh???

Marc GuntherMay 18th, 2012 at 7:06 pm

Really interesting post, thanks.

This is insightful:

Consumers "see increases in nominal earnings not as an effect of inflation, but as deserved rewards for their own skills and efforts. They then perceive any erosion of the purchasing power due to inflation as an undeserved loss."

I guess it won't be easy for politicians to turn to inflation to get us out of our fiscal mess–not if they want to be reelected!

kietgeneMay 18th, 2012 at 7:10 pm

India will soon see the downturn in next 1 or 2 years .
The politics are doing nothing either be India shining or dooming , they are fooling poor for petty votes and looting the public money so when the recession comes , they still can afford luxury.
The market cycle has been 1992-2000-2008-2016 ????

barfMay 19th, 2012 at 12:03 am

the only indicator of inflation that i am aware of is real estate. or what is now called "owner equivalent rents." Alan Greenspan "took care of that problem" in 2008 and we are all now left with "where do i have confidence in return of my capital" as interest rates are stuck on zero and prices plunge. That would be treasuries and EQUITIES and NOT real estate where without a doubt taxes are going to go up and where the cost of simply "holding" is exorbitant indeed. not true with treasuries or equities of course. the fact of the matter is the only thing keeping prices from totally collapsing and the economy plunging into "Depression 2.0" is the Pentagon and the Defense Budget…and ironically enough while we are all in agreement "the budget must be cut" it doesn't happen because of the peculiarities of politics in the USA." go figure…

dwbMay 20th, 2012 at 10:17 am

this post is an excellent illustration of the conundrum the fed faces: it cannot promise to hold down import prices which are determined on world markets. In fact, paying attention to such prices sends the wrong signal (if imports go up and therefore CPI goes up, and they tighten, they are compounding an AS shock with an AD shock). if import prices are too high, then people adjust by consuming less (i.e. fuel efficient cars – oil has gone up from $1 in 1988, but so has fuel efficiency and fuel efficiency has risen to the point where currently we are an exporter of some refined products).

Unfortunately, what your chart shows is that the political pressure to target oil prices is extremely high. But thats not a promise the Fed can keep, without causing substantial unemployment.

dwbMay 20th, 2012 at 5:50 pm

as always, thanks for reading comments (I am always pleasantly surprised when people take time from their busy schedule).

I wish i could take credit for the point, but a while back these guys names Bernanke and Gertler had a study saying that the reason high oil prices were recessionary was due to Fed macro response. but that was before the 2008 oil shock. hmmm.

I read Frankel's post (excellent), also Altig's rebuttal (Atlanta Fed macroblog) and (surprisingly) Altig posted my rebuttal of his rebuttal as a comment (personally i think Altig kinda cheated to cite Orphanides paper in favor of IT when the analysis is based on gdp deflator not PCE and when the results actually show the opposite under imperfect information. just my opinion).

Yes, i agree I think this is one of a very long litany of reasons to switch (back) to ngdp targeting. I did not mean to go there, but the Sumner corollary to Friedman's rule is something like "comments always and everywhere end up as an ngdp targeting phenomenon." seems like it, anyway.

Trent RockMay 20th, 2012 at 3:00 pm

Great article!!! I get the "have you BEEN to a grocery store lately?" a lot….There is also a HUGE myth that "the govt does NOT include food and energy" in the CPI…Or The BLS "does not 'report' food and energy in the CPI reports"…. another common error is the use of anecdotal evidence and hasty generalizations in determining price direction by Joe Schmoe Consumer..imho

PinewoldMay 20th, 2012 at 4:06 pm

These "reasoned arguments" are depressing. Look at medical costs, look at college costs, look at energy prices relative to income which has been nearly flat for years. Americans are slowly being crushed and you say energy prices are up less than last year!

lorenzofromozMay 22nd, 2012 at 4:24 am

Which is why Bernanke should talk about increasing income growth, as Scott Sumner suggests. After all, flat incomes make perceived price increases even more salient.

Mcalester WatercoolerMay 31st, 2012 at 5:44 pm

How come such discussions as this NEVER mention to so-called purpose of a 'free market' — the ability of the consumer to have choices on products and PRICES?

The consumer is complaining about the items that go up because he has no choice but to pay the price — no option. THAT the the REAL complaint — no choice.

For awhile the Internet has been giving some people choices — but the government (state and federal) are quietly doing away with that.

GuestMay 31st, 2012 at 6:04 pm

Interesting green line on the chart. Also interesting would be an average drawn through the green which appears to me would be much higher than the blue/red. And I keep waiting for a compensating drop in prices to compensate for the peaks. Big dips like the peaks never seem to occur. Curious isn't it.

idateasianDecember 23rd, 2012 at 10:06 pm

This is interesting feature for every one who wants to make their own website. Weebly web designer platform helps to create website with different categories as Business, personal, entertainment and many more.

George MeltonMay 5th, 2013 at 8:05 pm

How do you explain that gas prices doubled during Obama's first administratioon and food prices were up 10-25% yet CPI was basically flat. I would argue that the weightings applied to different components of the index do not match the buying patterns of the typical household.

Ed Dolan EdDolanMay 6th, 2013 at 12:05 pm

These are good questions. You can look up the information yourself on the FRED database published by the St. Louis Fed, or the BLS web site. Here are the numbers:

The CPI for all urban consumers rose by 9 percent from Jan 2009 to Jan 2013
The price index for food (weight = 7% of CPI) rose by 7% over that period
The price index for clothing (weight = 4%) rose by 7%
The price index for gasoline (weight=5%) rose by 106%

As you can see, some items rose faster and some slower than the average. Your own expenditure weights may be different from those of the average urban consumer. You should compare your detailed family budget with the averages. For example, if you use more than the average amount of gasoline, your personal CPI would rise faster than the average.