Why are observations of inflation so biased? And biased by gender?

by on November 16, 2012 at 5:32 am in Economics | Permalink

This is from an older piece, but I ran across it while browsing the internet.  It is from the Cleveland Fed, by Michael Bryan and Guhan Venkatu, 2001, excerpt:

The data indicate that the public’s estimates and predictions of inflation are significantly and systematically related to the demographic characteristics of the respondents. People with high incomes perceive and anticipate much less inflation than people with low incomes, married people less than singles, whites less than nonwhites, and middle-aged people less than young people. This Commentary describes what is perhaps the most curious observation of all: Even after we hold constant income, age, education, race, and marital status, men and women hold very different views on the rate at which prices are changing.

And now, more specifically:

In the roughly 20,000 responses we have received from our telephone survey since August 1998, the average rate at which respondents thought prices had risen over the previous 12 months was about 6.0 percent. This “perception” of inflation is more than twice the rise recorded by the Consumer Price Index (CPI) over the same period (2.7 percent). Further, if we separate our sample by gender, we find that the average inflation perceived by the nearly 8,500 men who answered our survey was 4.6 percent. While this response is higher than the official CPI inflation estimate, it pales in comparison to the 6.9 percent inflation perceived by the roughly 11,500 women who took our survey. What accounts for such a large discrepancy between the inflation rate perceived by the two sexes?

I found it through this piece, cited by Instapundit.  Is it possible that a high perception of inflation is largely the result of a relatively low real income, perhaps mixed in with a slight unwillingness to blame oneself for imperfect labor market prospects?  Does this help explain why tight money and stagnant median income have come together?

Saturos November 16, 2012 at 6:27 am

Maybe we should investigate how this correlates with differences in male and female consumption baskets… perhaps this explains the outrage over birth control coverage? And yet my impression is that that is still dirt cheap. So, I dunno, cosmetics?

Steve Sailer November 16, 2012 at 6:31 am

Higher income people tend to be smarter and smarter people tend to be more correct in their views, especially about quantitative questions. For example, a 2011 Gallup poll asked, “Just your best guess, what percentage of Americans today are gay or lesbian?”

Looking at the demographic crosstabs, it’s evident that low intelligence people were most likely to wildly overestimate the percentage of homosexuals: 53% of people making under $30,000 annually said that at least 25% of the population was gay, and 47% of those with no more than a high school education. 43% of Democrats versus 24% of Republicans got the question wildly wrong.

In general, people are terrible at estimating or remembering demographic statistics. A 2001 Gallup survey, right after the release of 2000 Census results, found that the average American estimated that 33% of the population was black and 29% were Hispanic. That adds up to 62%, but who’s counting? Not most people.

In that 2001 survey, nonwhites estimated that 40% of the population was black and 35% was Hispanic (adding up to 75%). In contrast, people claiming postgraduate degrees estimated that 25% were black and 24% Hispanic (only about double the Census numbers), which proves the value of advanced education.

Rahul November 16, 2012 at 6:56 am

Key difference between your example and CPI is that CPI itself is a somewhat artificial metric sensitive to what basket is chosen. Maybe we should be criticizing the metric more and not the respondents.

If I had a hotel where my consultants calculated a comfort-score which was always very high and yet customer-surveys revealed that everyone was complaining about uncomfortable rooms; would I be ignoring my “stupid customers”?

Especially for inflation it is the perception that matters, and if our measures don’t capture it well, change the measures.

Corporate Serf November 16, 2012 at 10:00 am

To expand on this point, richer people probably have consumption baskets closer to the actual CPI consumption basket. Also, for poorer people the correct perception might be to not de-season the personal cpi series. What season was the survey taken in?

The situation with the hotel is not quite the same. The people who calculate and use the CPI measures don’t face any adverse impact if the measure is off. It is not a competitive private market.

Rahul November 16, 2012 at 10:29 am

Agreed. I don’t see many comments about validating the basket. Maybe we’ve become such a heterogeneous population that a single inflation value doesn’t make any sense:it represents no cohort.

On the other hand the systematic skew is distressing. Men and women and other sub-populations might have a relatively different impression of inflation; yet in all cases the perceived inflation is higher than the CPI inflation.

To me that’s a strong case for reassessing how we calculate that metric. The basket of commodities doesn’t make sense.

Claudia November 16, 2012 at 10:45 am

Now let’s not throw the baby out with the basket. The question of interest is going to determine the most useful measure. Households spend different proportions of their income on the same goods/services AND some households spend a lot more than others in total. Measures of aggregate prices changes capture aggregate patterns and not necessarily the experiences of a typical household. You have to be very careful when comparing across aggregate and household measures (not to mention very different survey methodologies). Each approach has value, but one should not negate the usefulness of the other.

Rahul November 16, 2012 at 10:53 am

@Claudia:

“Measures of aggregate prices changes capture aggregate patterns and not necessarily the experiences of a typical household. ”

Agreed. But how do you explain the skew? What’s the quality of the aggregate metric if it is systematically and significantly above or below the population average.

Unless this study is wrong or an outlier. Which could be.

Michael Cain November 16, 2012 at 11:24 am

“But how do you explain the skew?”

Poor households spend much larger fractions of their income on energy, food, and health care, all of which have seen price increases higher than the CPI. IIRC, the Census Bureau found that households at the 90th percentile spend 3% of household income on energy, a median household about 6%, and a household at the 10th percentile spend about 16%. Gas prices tripling from 2000 to 2012 — although the path was far from smooth — hit the poor household much harder.

Brian Donohue November 16, 2012 at 11:26 am

Have you consdered widespread innumeracy and general obliviousness? How much different is this than the inability to locate the Civil War within a 50-year range, except that nobody’s gonna plump for negative inflation, so the results appear to be skewed high?

Corporate Serf November 16, 2012 at 11:49 am

@Claudia

I think pervasive computation might change things qualitatively. We might not need a “global” CPI measure if it is easy to create personalized measures. You just need to pull your credit card purchase records. My amex card statement breaks it out in buckets….

Perhaps, it is a function of my profession, but I am seeing traditional economic modeling giving way to individual people being able to get the tools to make the right decisions for their circumstances, and, hopefully, we will see less of “Clearly poorer people make poorer estimates of CPI because they are stupider (paraphrased)”

Claudia November 16, 2012 at 4:26 pm

CS, the mandate of price stability applies to the aggregate not each individual. Summary, aggregate measures of inflation are always going to be needed. That said technology like you described may greatly improve measurement and maybe even awareness of price changes at the individual level which feeds into those aggregates.

The Original D November 16, 2012 at 6:49 pm

I don’t know how much you can expect people to know. Most of us are low-information about practically everything, so we look to experts to guide us.

Knowing the exact actual inflation rate is much less important than whether our financial advisor knows it.

Corporate Serf November 16, 2012 at 8:54 pm

@Claudia,

There seems to be several uses of the CPI in “official” calculations: COLA (as someone mentioned), TIPS bond adjustments, others?

All of these seem to involve things the federal government pays out. So there is an incentive on the Government’s part tp underreport inflation. Plus getting the number lower is also good from the perspective of how it looks.

Now federal employees are pretty powerful interest group. Perhaps they weren’t as powerful in the nineties when the “hedonic adjustment” to legally underreport inflation was phased in? Or were there some other quid pro quos? The TIPS market is pretty thin and I think buyers of those bonds understand that the government itself is reporting the inflators, so there is a conflict of interest.

Years ago (when Greenspan was still considered a good guy among liberals), Greenspan was on the Jon Stewart show. Stewart asked one good question: if we believe in free markets and everything, why does the feds set the overnight interest rate? Greenspan smiled and said that was a great question and the great Stewart, the slayer of the Cramer, simpered like a besotted schoolgirl. So, if a quasi-government department has a “mandate” for price stability, that would make the CPI a great number all of a sudden? Also see Libor vs OIS. India had these wonderful government departments perpetually promising us great things in the next “5 year plan”? Nothing changed until the government ran out of money and was forced to get out of the markets by the IMF. Too bad the IMF is not strong enough to force the developed country governments out of the credit, rates and FX markets.

Bottom line, summary aggregate numbers are needed only if the government wants to meddle in the market, otherwise not.

tgrass November 16, 2012 at 10:41 am

My understanding of the CPI is that is designed for COLA for federal payments. If I remember correctly, under Clinton the basket was revised to reduce the COLA. The CPI would then have HUGE adverse impacts if it’s “off.”

Mike November 16, 2012 at 12:13 pm

They added things like substitution effects as well as quality adjustments. For example, if the price of beef goes up, I’ll buy more pork and less beef. Also, a Toyota Camry today offers a lot more in terms of comfort and safety than one from the 1980s, so comparing it purely on price leaves out the gains consumers get in quality.

I can’t speak for the precise methods involved in these adjustments, but ignoring these factors made the old CPI clearly overstate inflation.

gwern November 16, 2012 at 10:35 am

> Key difference between your example and CPI is that CPI itself is a somewhat artificial metric sensitive to what basket is chosen. Maybe we should be criticizing the metric more and not the respondents.

When the difference between perception and measurement is >100%, this suggestion will only pass muster for a little while before the exponential growth in difference becomes a little absurd. And I don’t see anything in the paper to suggest that this is not a long-standing differene.

(So apparently inflation is a bit like a Shephard tone or barberpole – it just keeps going up and up and up…)

Greg G November 16, 2012 at 9:03 am

There is a general loss aversion bias that causes most people to be more impressed by losing something than by gaining something else of equal value. For this reason, the many things whose real prices have gone down leave less of an impression than the things whose prices have gone up.

If you can afford everything you need this will make less of an impression than if you actually have to stop consuming something you used to enjoy. So it is not a form of stupidity that causes low income people to feel that inflation affects their standard of living more than it does for wealthier people.

rpenm November 16, 2012 at 9:16 am

+1 to Greg G.

Mike M November 16, 2012 at 9:57 am

+2 to Greg G.

Case in point: I was talking to my father in law a few weeks ago and in two consecutive sentences he was able to say (1) the blue-ray player he bought 5 years ago was $900 but today it goes for $99 and (2) inflation is out of control – just look at gas prices.

Richard November 16, 2012 at 10:02 am

-1 to Greg G.

The question was not “how much does inflation affect you.” The question was “what’s the inflation rate.” The inflation rate is something published in newspapers. This is a test of how well-informed people are, pure and simple.

Poor people read less than wealthy people, which is probably why their answers were more inaccurate.

Women read more fiction than men, who read more non-fiction. Again, reading patterns explain the results.

sort_of_knowledgable November 16, 2012 at 10:48 am

Looking at the paper the question was not “what’s the inflation rate.” but
“by what percentage have prices increased or decreased over the past 12 months?” and notes “Since we do not yet fully understand how people are interpreting survey questions on inflation, we cannot know what benchmark their responses should be judged against.” It is not meant to be a test of how well people were informed of official government statistics.

The Original D November 16, 2012 at 6:53 pm

Yes, but the only reason anyone besides economists cares about inflation is it’s personal effect. It’s not the same as asking “What is Planck’s constant?”

Dent November 16, 2012 at 3:54 pm

+6.02*10^23 to Greg G.

Noumenon72 November 17, 2012 at 6:49 am

You’re giving him a mole?

Brian Donohue November 16, 2012 at 10:15 am

Perhaps loss aversion is not stupid. Perhaps.

It is often understandable and explainable, thanks to Kaheman et al.

But it’s not rational, optimal, or desirable for those making decisions.

If all you want to do is explain and defend why people make the choices they do, your work is done.

If you want to help people make better, more rational, optimal and desirable choices, there’s a lot of work to be done. To me, this is where behavioral economics is interesting and useful.

Corporate Serf November 16, 2012 at 9:05 pm

Doesn’t the “optimal” depend on the “Cost Function” (or, as an economist would say, the “Utility Function”), and the same set of constraints on the critical path? So do you propose to force everyone have the same cost function and basically the same set of constraints on the critical path?

You can see where I am coming from. I don’t want to tell others what decisions to make (except my kids and that too less and less) and I don’t want others to tell me what decisions to make. I don’t want to help people make better, more rational, optimal, desirable (desirable for whom?) choices. I don’t want to explain, nor defend why people make the choices they do or don’t.

Brian Donohue November 16, 2012 at 11:40 pm

This sounds like a knee-jerk reaction. Where did I say anything about forcing anyone to do anything?

Kahneman has plenty of examples of framing the identical decision in different ways and eliciting different choices. Some of these examples include very smart people, like doctors. Is this rational? When people consider multiple, different frames for decision-making, they often make decisions that they think are better for them. I know I do.

I got no problem with you covering your ears and doing whatever the hell you want though. Be my guest.

Floccina November 16, 2012 at 11:40 am

Reminds me of a funny story. I knew a guy who grew up in a predominantly black part town and he said he was shocked when he found out there were more whites than blacks in the USA and being a kid at the time, he thought “why don’t we rise up against them and stop them from pushing us around”.

jcw46 November 17, 2012 at 10:26 am

Maybe it’s what they’re seeing on TV?

In an effort to be seen as inclusive, many shows over represent many minorities if you look at the industry overall. Also certain sports and therefore the color commentators have a larger black component.

Any one who claims that TV doesn’t have an influence on how people think, what people think and how they perceive the world is lying. For good or for ill, TV has a disparate impact on the culture let alone consumer trends.

Time for some regulation of the potential for manipulation and the transmittal to large segments of the population of inaccurate information.

Trouble is, I wouldn’t trust the industry to police itself nor a governmental agency and neither would most of the people. So it will be let continue as it is.

Steve Sailer November 16, 2012 at 6:33 am

Second, the CPI _does_ underestimate the rate of inflation in the cost of buying a home in a good public school district, which is, increasingly, the entry price for marriage and legitimate children.

Slocum November 16, 2012 at 7:27 am

In that domain, we’ve had a great deal of deflation — prices have fallen dramatically and mortgage rates are at historic lows.

RAR' November 16, 2012 at 8:33 am

Lately, yes, but this data is from before the housing bust.

Thomas November 16, 2012 at 9:09 am

CPI focuses on rental equivalence value, and so there’s only indirect effect from falling prices, mortgage rates.

Steve Sailer November 16, 2012 at 6:38 am

Judging by the gender balance of real estate agents, women are more interested in the cost of homes, while men are more interested in computers and cars, both of which have not seen much inflation.

In general, though, women on average are not as well informed about serious current events as men are, so it’s hardly surprising that women are more wrong about quantitative public affairs questions:

http://anepigone.blogspot.com/2011/11/old-republican-white-men-know-whats.html

Nick November 16, 2012 at 7:13 am

Isn’t there a general tendency for women to do the shopping still (not saying this is ideal, and happy to have my sterotyping proved wrong). Given food prices have increased much more quickly than other prices (and non-food purchases are more sporadic/less visible) this result isn’t that surprising. Some form of behavioural anchoring…

Also, the CPI adjusts for the quality of the goods whereas I imagine people dont do so mentally, though this is only an explanation for the divergence between all people and the CPI reading and not the gender gap.

RAR' November 16, 2012 at 8:42 am

This, and also what Ted Craig says below about food and fuel. If you assume the stereotype is true that most women do the grocery shopping (in my house, that’s not the case, but we’ll assume it anyway), then their perception of inflation may be closer to a poor person’s perception of inflation because both see the price of food (which is not including in CPI) going up. For the poor, food prices may skew their perception of inflation more than it does wealthy women because wealthy women can afford more than just the basics and so have more samples to go by.

Of course, people are generally bad at estimating such things. But explaining why everyone overestimates inflation and why some people overestimate more than others are two different things.

zbicyclist November 16, 2012 at 10:03 am

The price of food is included in the CPI.

Mo November 16, 2012 at 11:24 am

But if a person’s basket (or if the portion of the basket they see) is disproportionately food, they would think inflation is higher than it is. If, purely hypothetically, my wife did all of the shopping except for electronics and I all I purchased was electronics, my wife would have a pretty accurate view of inflation and I would perceive a deflationary environment. This is despite the fact that electronics is in the basket.

AN_PD November 16, 2012 at 9:02 am

I would agree with this. Even though there are increasing numbers of households in which men do the grocery shopping, it is still the case in the vast number of cases that women do most of such shopping and therefore our perception is disproportionally affected by increases in the cost of food. More generally, women tend to be responsible for other day-to-day type purchasing (clothes, school supplies) and are, generally speaking, responsible for making sure that stagnant or shrinking household wages stretch to cover the household’s needs. Men are more likely to be involved in big-ticket items (mortgage, car, major appliances) and the infrequency of such purchases may make men less in tune with price increases. Correspondingly, women are more sensitive to the gap between household income and prices and may perceive more inflation as a result.

Nick November 16, 2012 at 7:16 am

“Is it possible that a high perception of inflation is largely the result of a relatively low real income?”

Another way of putting it would be in sociological terms: fear. All of the mentioned groups that have higher inflation perception are also the lesser-advantaged people in society, who have a relatively harder time finding jobs, tend to have less income, and are also the first to get the boot when recessions arise. The instability means that inflation has a bigger impact on those groups than the high-incomes, the whites, and the married, who tend to have more stable prospects. Because it’s relatively more concerning, they also fear it more, which leads to larger estimates as a part of being more cautious about it. The same thing happens with any number of issues where one side fears something more than the other: Republicans fear attacks from abroad, and so perceive more hostile foreigners (and indeed of foreigners as hostile) more often than Democrats, and so on.

Cliff November 16, 2012 at 10:23 am

But we are talking about men vs. women and controlling for all those other things.

Nick November 16, 2012 at 6:14 pm

I know, but regardless of how well off you are individually, people still identify with their groups: male, female, white, black, hispanic, rich upbringing, poor upbringing. As Jennifer Lopez sings about in her informative discourse on cultural psychology, she’s “still Jenny from the block.” Remember, perceptions aren’t always (sometimes, aren’t often) based on objective reality, but how people *feel* about the world, which includes all of the cultural baggage from their identifying groups. If women as a whole are taught to be more careful and cautious, then an individual woman may still be careful and cautious about money even if she’s personally high-income.

Ted Craig November 16, 2012 at 7:25 am

Isn’t it that the two factors taken out for core inflation – food and fuel – consume a larger portion of poor people’s income?

IVV November 16, 2012 at 7:30 am

I’d agree that whoever does the grocery shopping will see higher inflation than whoever does not. Furthermore, it’s important to recognize that for the general public, “inflation” means a rise in consumer prices. With the nonexistent wage inflation experienced by the public, raises don’t count.

So, if the price of milk goes up by a dollar, but the price of everything else remains the same, then an economist would observe that there is very little inflation, but grocery shoppers would see a whole heck of a lot of inflation. In addition, lower income brackets would see more inflation than higher income brackets, families would see more than singles who see more than couples without children, and so on.

What we’re seeing here is a disconnect between the across-the-board inflation the economist measures and the drop in real discretionary spending the consumer experiences.

JAK November 16, 2012 at 7:41 am

What is the actual question that the survey asked?

Inflation for me is change in prices of the stuff I buy. So instead of interpreting the survey as some groups overestimating the inflation, we can probably draw some indications on how the actual price changes are impacting various groups. Getting such input to better target the tuning of policies is much better outcome than on perpetuating the basket case focus of inflation measurement. All pun intended.

Patricia Mathews November 16, 2012 at 7:46 am

Women do the grocery shopping, women get the car filled, and women have to get the kids’ school uniforms and supplies. Women write the checks for the utilities and the insurance and health care.The price of all those things has gone up. The price of the toys of affluent men has not..

And in no other economic area do people blow off the purchasing manager than this … but blow her off you all apparently do, even today. “Could it be …. cosmetics?”

Well, that, too, especially if her managers demand “a professional appearance” including makeup and coiffure. Which believe me, some do, adding to her operating expenses.

Ron Potato November 17, 2012 at 12:25 pm

Women pay the bills and only the men buy toys? What planet are you on?

doctorpat November 18, 2012 at 9:38 pm

What planet are you on and how do I move there?

Heath White November 16, 2012 at 7:58 am

I think if you code “how high is inflation?” as “how financially insecure do you feel?” you get a pretty good explanation of the results.

Cliff November 16, 2012 at 10:26 am

Then the question remains, why are women more financially insecure than men when holding income equal?

Michael Foody November 16, 2012 at 8:03 am

Prices generally increase, but inflation is really a blend of prices on certain things going up a lot, prices on certain things going up a little, prices on certain things staying the same, and prices on certain things falling. Large changes in either direction are likely to be more noticeable than prices staying the same. Since prices are rising generally this would tend to lead people of either gender to overestimate the rate of inflation (indeed they do). Shopping for a one off product does not really give one a sense of price changes, if you buy a desk or a sweater as a one off you are unlikely to be able to make meaningful inferences on the movement of prices. Observations of inflation are most likely when you are buying the same product over an over again. This is a big part of why the price of gas looms so large in people’s minds. Grocery shopping is the other type of shopping that entails repeatedly buying the same goods. This is a task that routinely falls to women rather than men.

Based on this theory I would predict: That men who do grocery shopping would report inflation estimates higher than men who do not. I think what’s going on is more observations, filtered through cognitive bias, actually leads to inferior estimates.

Brian Donohue November 16, 2012 at 8:16 am

But I thought Republicans were stupid.

Disclaimer- I’m white, middle aged, married, high income, and male, and I’ve been wrong on inflation in the past several years, so at least I’m stupid. But I’m not really Repbulican, at least according to most Republicans I know.

bellisaurius November 16, 2012 at 8:19 am

Me and the wife shop together most of the time. I’ve noticed rises in some of the more variable items like meat and fresh vegetables, but some of the rise gets hidden by changes in package sizing, but more importantly, and I say this since my laziness means I have a sampling of receipts I see from time to time, that i substitute like crazy as I treat it like a competition to get the most bang for the buck, so my bill is almost always between a 100 to 150 dollars or so.

Ritwik November 16, 2012 at 8:54 am

Inflation being conflated with the price level, I think. Stagnant real incomes could be chief explanation.

DocMerlin November 16, 2012 at 8:59 am

Simpler explanation: Its not biased.
They are each observing inflation they are unable to arbitrage away from in the basket of goods they buy.
The wealthy are more able to protect themselves from inflation so they experience less.
Women are more likely to be the family’s grocery purchaser. They buy the goods most sensitive to inflation.
etc.

John Thacker November 16, 2012 at 10:17 am

Possible, but I feel that groceries are actually less sensitive to fluctuations than gasoline. With groceries, it’s a lot easier for people to substitute various goods and buy what’s on sale. Gasoline there’s much less of an option.

Anthony November 17, 2012 at 10:22 pm

If you switch from beef to pork because the price of beef has risen, you don’t think that prices haven’t risen because you’re spending the same amount on meat.

Also, the CPI adjustments in the 80s and 90s may have switched from signififcantly over-weighting housing cost increases to significantly underweighting them. See, for example, http://www.shadowstats.com/alternate_data/inflation-charts

Claudia November 16, 2012 at 9:20 am

I once asked someone who has run a household survey for decades what still puzzles him in the data. One thing he said was the pervasive gender differences in the responses. We were talking broadly about the survey which includes a lot questions about economic conditions, including inflation. I doubt the differences are all about who buys what in the household or even across households. Maybe it’s living in different worlds or filtering information from the same world differently.

More to the post…if you’re interested in learning more about survey measures of inflation I would recommend the recent work by Wilbert van der Klauuw at the NY Fed and co-authors. They have done much work to unpack inflation measures and even how inflation expectations affect behavior. Not sure I support all their methods or conclusions, but it’s really good stuff.

Brian November 16, 2012 at 10:16 am

So men said 4.6% and women said 6.9%. Wouldn’t it make more sense to say there were many people guessing around 5%, but more women guessing 7.5% or 10%? How was the survey conducted?

Eddie November 16, 2012 at 10:57 am

I own a house. My perception of inflation will likely be muted by the fact that 30% of my spending is not subject to inflationary pressures. I am less concerned about a 20 cent rise in the cost of eggs as long as the thousands I spend on my mortgage is not going to increase. I can eat less (or cheaper) eggs however, my mortgage is (for all intensive purposes) a constant.

KnowPD November 16, 2012 at 11:13 am

Is this an argument for NGDP targeting? If inflation is systematically over-estimated by some groups, would “growth” be over-estimated also?

Matthew C. November 16, 2012 at 11:32 am

Perhaps if CPI weren’t specifically designed to understate inflation (to make growth look better, and control government entitlement expenditures / COLAs), the CPI numbers would be much more inline with actual experience.

Also “inflation” is very problematic metric — the bottom line is that monetary dilution (ie: money printing) is a great mechanism for screwing the poor and middle class in the interests of the wealthy, whose income streams and assets are bid up by all the monetary liquidity.

collin November 16, 2012 at 11:39 am

Considering this is 11 years old and the study stock includes a lot of housewives from the 1960s and 1970s I suspect some of this difference has diminished. I suspect the main driver of inflation rates is various food prices and decades ago women did a lot more food shopping.

CR

FE November 16, 2012 at 11:53 am

What about differences in who pays attention to financial news? I doubt anyone can estimate inflation very accurately from their own consumption. I think the question as getting at what kinds of people have heard the official number often enough to anchor their estimate around it.

Vernunft November 16, 2012 at 11:55 am

Biased by how feminine or masculine people are? Not by sex? Odd.

Adam Dynes November 16, 2012 at 1:10 pm

Political scientists have found that views on inflation and economic performance vary by partisanship. When your party is in power, you’re much more likely to say that inflation is lower and economic conditions are better overall.

This is often blamed selective exposure (e.g., liberals watching MSNBC, conservatives watching FOX News) and partisan filtering (i.e., partisans filter out information that doesn’t comport with their partisan beliefs).

There’s some new research that argues that a lot of it is partisan cheerleading, especially since there’s no cost for cheerleading on surveys. If you pay people even less than 50 cents per correct answer, you can cut the partisan difference on many of these questions in half.

jdn November 16, 2012 at 3:06 pm

Why would you expect people to accurately estimate inflation at all? You’re expecting people to basically determine the change of price in probably thousands of items over the course of a year, know what is in the official basket of goods that make up the CPI, and calculate the inflation rate. The idea that the answer to that question is some sort of “observation” is silly.

Really, what they’re asking people people to do, is to make up a small percentage number based on their feeling about the economy. The fact that different groups come up with different numbers may be useful for a parlor trick, but that’s about it.

Jennifer November 16, 2012 at 3:14 pm

If you are interested in how different socio economic groups might experience inflation, the Australian Bureau of Statistics creates several subindices of inflation to assist policymakers in adjusting different benefit types. You can see the details here: http://www.abs.gov.au/ausstats/abs@.nsf/mf/6467.0. For example, in the year to September 2012, the average CPI (basket of goods reflecting the population as a whole) was 2.0%, which was the same as the inflation experienced by age pensionerse. The inflation experienced by households headed by employees, however, was 1.0%, as their basket of goods was different.

So richer people thinking inflation was lower is probably a combination of actual experience (as their consumption is weighted towards lower inflation purchases) and perception (as several commenters have noted above).

Willitts November 16, 2012 at 6:04 pm

Their perceptions are biased for the same reason any estimator is biased – their observed sample isn’t random.

A person’s market basket won’t come anywhere close to matching thr national market basket, and what a person buys or more important to them than what people buy in general.

Elasticity of demand and hence perception of inflation is based on prices relative to budget. So obviously the rich have lower perceived inflation.

But this isn’t perceived – it is individualized actual measures of inflation. If all of these were averaged, it should be close to the national inflation rate.

There are distributional consequences in basing monetary policy on the national CPI. I’m not suggesting the Fed should necessarily care about these consequences, but they shouldn’t be surprised to hear a lot of grumbling. This is when the drunk at the corner bar understands economics better than Ben Bernanke.

jorod November 16, 2012 at 8:20 pm

That half gallon of OJ is now 59 ozs. but the price is the same.

alwaysfiredup November 17, 2012 at 3:31 am

Does no one else notice that the divide mirrors our political divide? Looks like conservative people are prone to estimate conservatively and liberals prone to estimate liberally.

Brian Otto November 18, 2012 at 11:16 am

First, let us no assume that the estimates are wrong.
Different people may be buying different things that are rising in price at different rates.
So, as opposed to the guy (Steve Sailor) who blithely assumes that smarter people tend to have
greater incomes and thus attributes the overestimate of inflation by poorer people to the
mistakes ‘low intelligence people’ make, we can actually differentiate between the reality
of a high income person and a low income person. Sailor mentioned that his purported
‘low intelligence people’ wildly overestimated the number of blacks and hispanics in the population, but
guess what? Most of those people might just live in neighborhoods that are 60 or 70% or more
black and hispanic. Mr. Sailor may know his demographics, but he does not seem to recognize that we
live in a segregated country and most of those people were not misestimating because of their
‘low intelligence’, but making a reasonable guess based on their daily experience.
As a teacher in the Bronx, I used to take my black and hispanic students to
majority white areas and they were amazed that you could be somewhere and not see a black or
brown face. So once we factor in that people have different experiences, then maybe their guess
are closer to reality than the CPI is. Maybe men buy goods that are less likely to go up in price, maybe
women are more likely to do grocery shopping and are particularly sensitive to hikes in prices. Maybe
poorer people are also sensitive because price hikes eat up a much greater portion of their discretionary
income while richer people either draw on savings, don’t buy as much or leave a bit of a balance
on their credit cards.
I do think there is more to it than that — people probably make such estimates based on a couple of price
changes and then extrapolate. The reason that higher income people are closer in their estimates is not,
I don’t think, because they are of superior intelligence — although for Mr. Sailor and his ilk that would help to
justify inequity as a just outcome — but because of their choice in reading and viewing materials. A higher
percentage probably read business columns or news shows that would report the CPI Indeed, it may be that they are not making an estimate at all, but merely reporting what they heard on the radio.

Comments on this entry are closed.

Previous post:

Next post: