The consumer hourglass theory?

P&G isn’t the only company adjusting its business. A wide swath of American companies is convinced that the consumer market is bifurcating into high and low ends and eroding in the middle. They have begun to alter the way they research, develop and market their products.

Food giant H.J. Heinz Co., for example, is developing more products at lower price ranges. Luxury retailer Saks Inc. is bolstering its high-end apparel and accessories because its wealthiest customers—not those drawn to entry-level items—are driving the chain’s growth.

Citigroup calls the phenomenon the “Consumer Hourglass Theory” and since 2009 has urged investors to focus on companies best positioned to cater to the highest-income and lowest-income consumers. It created an index of 25 companies, including Estée Lauder Cos. and Saks at the top of the hourglass and Family Dollar Stores Inc. and Kellogg Co. at the bottom. The index posted a 56.5% return for investors from its inception on Dec. 10, 2009, through Sept. 1, 2011. Over the same period, the Dow Jones Industrial Average returned 11%.

“Companies have thought that if you’re in the middle, you’re safe,” says Citigroup analyst Deborah Weinswig. “But that’s not where the consumer is any more—the consumer hourglass is more pronounced now than ever.”

…Firms catering to low-income consumers, such as Dollar General Corp., also are posting gains, boosted by formerly middle-class families facing shrunken budgets. Dollar stores garnered steady sales increases in recent years, easily outpacing mainstream counterparts like Target Corp. and Wal-Mart Stores Inc., which typically are more expensive.

Here is more, no need to click on their silly links.


The middle-class* in the western world will be remembered as a Twentieth Century anomalous phenomenon.

*The American English sense of the concept, not the British English sense.

"Middle class" is one of the most brilliant newspeakisms in history. If everyone thinks they're "middle class" it's impossible for people to even discuss class problems. Then the rich win.

There have been a few attempts to impose systems in which the rich didn't win. How well did that work out?

The United States had big taxes and big gov't 50 years ago and we had the #1 economy in the world. Funny how cooperation and sharing and redistribution in an economy can lead to stability, shared prosperity and global advantage.

So I'd say it worked out pretty well. Too bad we're ditching the American system for the Russian one. How is that working out?

Yes, we had the No. 1 economy 50 years ago. We also had the No. 1 economy 100 years ago without hardly any income tax. I'm not arguing taxes are bad (although I'm not a fan), but correlation isn't causation. Also, it's a little easier to be No. 1 when the rest of the world was reduced to rubble just a few years before.

Fifty years ago, the rich were still winning and people were still complaining about it.

Really, it doesn't matter how much the everyone's standard of living goes up. As long as "the rich" exist, there is a problem. This, to me, is exemplified by Tomasz's comment suggesting that "middle class" is some sort of Orwellian word invented to keep poor people down--as if the widespread, unprecedented prosperity that the term describes was not a real phenomenon.


There is an interesting "Philosophy Bites" podcast about this. Alex Voorhoeve argues that in some contexts gross inequality by itself is evil even if everyone has been rendered better off.

One of the arguments is that inequality is corrosive to the social fabric and he offers some other arguments too. Caveat emptor.

Rahul, thanks for the interesting link.

50 years ago, we were the only industrial power of significant size still standing after WWII with most of Europe and Japan lying in rubble.

Sbard: that's the Broken Windows fallacy on a planetary scale (and happens to be false besides). Also, why did US prosperity persist for years after Europe and Japan had rebuilt (with US funding!) and were strong, competitive ecobnomies in their own rights?

Does the French Revolution count?

@JonF "that’s the Broken Windows fallacy "

No, it's not. The BWF is a statement that the replacement of destroyed wealth creates wealth. Sbard's is saying that the US was the only power whose wealth wasn't destroyed, a quite different (and largely true) idea. Contemplate this on the Teufelsberg. It also happens that the US was the only country whose wealth creation capacity was also largely intact, including both the physical and social infrastructure. This led to all kinds of wildly wrong theories, including the belief that the GM corporate structure was the zenith of efficiency. You'll note that US leadership in most areas persisted until the 1970s, about the time that the middle-class started seeing fractures.

In societies that measure success at least to a good extent in terms of wealth, the rich have won by definition. I think what most people are concerned with has more to do with class mobility, which has been pretty high at least since the end of the 18th century. But with liberalism working just as advertised since then, the long term effect clearly is a new equlibrium with class by mental capacity instead of birth. Being to a high degree genetically determined, class mobility has thus come more and more to a halt.

Even assuming perfect class mobility and perfectly equal opportunities, society would not be organized by mental capacity. There are many other factors, such as work ethic, individual values and preferences, social skills, physical appearance, etc.

Oh, no doubt about that. But work ethic, values and preferences, even physical appearance are all positively correlated with IQ.

I wonder if this is a special theory so much as it may simply be empirical evidence for the 'disappearance of the middle class' that liberals have been fretting about for awhile and conservatives have been ignoring because it's kind of annoying . (Still, at least 99.6% of poor households have refrigerators and all that.)

P.S. You are free to remove the silly links from your own text.

As a conservative, I do find it kind of annoying--that people keep asserting this without arguing for it. Here are U.S. census data:
The inflation-adjusted income of every segment of the American population has consistently grown over time. Perhaps you can argue that some of this growth is not real, for whatever reason. Perhaps you think it is not fair the income of the rich has grown faster than everyone else's income. But to claim that the middle class has "disappeared" based on these data is counterintuitive, to say the least.

"Disappearance of the middle class" phenomenon refers not to the fact that middle class has already disappeared, but to the fact that we are on a trajectory where all or most of recent gains in GDP went to the rich not to the middle class. If this trend persists, then middle class will eventually disappear. I think a good counterargument would be that there is no law of physics that guarantees that it is possible to ensure that gains always get redistributed equally. There bound to be times when only the most industrious are able to capture all the gains and it's doubtful that can be fixed without lowering everyone's standard of living. Having said all that, the trend is still worrisome, no use denying it.

Well, even if the gains continue to be distributed in this fashion a 'middle class' will still exist since we would still have a group of people who are not rich but are not poor. The only difference would be that the rich would be richer... The argument for the middle class disappearing would be valid if that group's income was decreasing.

What is the probability that what we're really seeing is nothing more than a demographic phenomenon? One where the "rich" are really the members of the very large Baby Boom generation who collectively are all at their peak earning years, and the "poor" are the very large "Generation Y" generation, who are all collectively at their years of lowest earning power, because they're still entering the U.S. workforce. The middle class then would be represented by the much smaller "Generation X".

Taking those changes into account, what the companies are doing then is reshaping their marketing to fit the changing age-based population distribution of the United States.

As for the "consumer hourglass theory" - if that theory is representative of the kind of analysis that Citigroup produces, where they don't really appear to understand what's really behind it, is it any wonder that they've become such a basket case?

Ironman --- that is an interesting thesis, but if you do panel comparisons (ie look only at 35 year old male workers from the mid-70s, mid-80s, mid-90s, mid-00s) the trend is for lower inflation adjusted incomes among men entering their peak earning years.

The disappearing middle class factoid gets thrown about a lot but I am confused as to where the data shows this:

e.g. Here is one histogram of American household earnings (admittedly a bit dated; circa 2005).

Where are the rich and poor spikes? The middle class seems strong enough.

Heheh, agreed - that graph says a lot.

I suspect people making 50-100k feel put upon because that's where government help phases out and taxes phase in, but that's just a guess. It implies that looking at household cash outflows would be more informative than looking at inflows.

I suspect it's not so much an income trend, rather a spending trend. Investment analysts really don't care about an average individual's income, but care quite a bit about where that income is spent. What they're really noticing is that more and more of the people whose incomes are in the middle of the pack increasingly don't differentiate between the cheapest good in a category and all goods except the luxury option, so for some items they buy very cheap, for others they buy luxury but less is spent on items that aren't at one end of the spectrum or the other (you have to realize that Wall St defines not failing as revnue growing by more than 4%/yr and profits growing by 2-3x that). Products that used to compete very successfully with products aiming for the middle of the market, aren't growing anymore.

Bingo. Income may play a part, but spending patterns are more important. I think a lot of people (at almost all levels of the income spectrum, perhaps barring the lowest and highest extremes) have come to the realization that, at least these days (this may not have been true in the past), there's bugger-all difference in actual quality and performance between the middle-of-the-road-recognized national brand and the cheaper house-brand version; so for most of their expenditure, they will buy the cheapo version (say, when it comes to noodles or t-shirts), and set aside a large chunk of income for a few items (say, an iPad or a hotel stay) where they can either "treat themselves" or impress their friends.

I believe this bifurcation has been well documented in the Economist and elsewhere.

@nelsonal & kiwi dave

What you say further convinces me that the "disappearing middle class" theory is a bit weak.

I wouldn't go that far -- I think there is some indication that there is a shrinking middle class (in the sense that the word is used in America, rather than Britain) in many developed countries, but the bifurcation of markets for consumer goods is probably not an example of this.

If you draw a line on that histogram to separate low from middle income, by whatever criteria (as long as its in real dollars), I bet you'll find it moves steadily to the right with time after 1970 (TGS, anyone?). Considering the steepness of the drop, that results in a rapidly declining middle.

Furthermore, if you made the histogram in terms of wealth (counting dollars instead of people), I bet you would find the high-income end growing rapidly. That is, the number of wealthy people is not increasing so much, but their share of the wealth is.

Well, from that graph, a family of four where the two parents each make $25k a year is middle class (that's roughly the median household income..actually, in 2005, it was $46k, but close enough). I think that given the cost of health care, education, etc. people making $25k each don't feel particularly "middle class" as compared to the "middle class" lifestyle of the mid 20th century.

Watch Fox News and you hear about how people who make $25k a year need to pay more taxes. Meanwhile the channel whines like a petulant child whenever someone suggests that people in the top 2% of the income distribution should be considered "rich."

So what is the middle class? The 40% of households that make between $36 and $90k a year? aka families where the husband and wife each make, on average $18-45k a year? Neither my wife or I never made as little as $18k a year, but even with that, we'd still be making more than 40% of the country, which is pretty darn close to "middle class." The fact that one can even consider an individual making $18k a year as "middle class" I think is what gets people riled up about the whole issue.

If you want numbers for families of 4, see

Median household income for a family of 4 in 2005 was $70,312.

The overall household income statistic is skewed by the fact that, again as of 2005, 27% of households have only one person and another 33% have two people. See

$46k for a single person can be a pretty decent living. For two people I agree that it can be a stretch, strongly depending on cost of living.

$70k for a family of 4 is not amazing either, of course.

One other important thing: households include students and the retired. The latter typically have slightly income than expenses: that's what retirement savings are for. The former don't have much in the way of income. So having higher income than 40% of households is not that meaningful a statistic.

Last thought. Would you consider a student who's making $18k a year in a part-time job, studying mechanical engineering, and whose parents are a teacher and a nurse middle-class? Because I sure would! The issue is that people think of the aggregate numbers as applying to people in their prime earning years, but for those the income distribution almost certainly looks very different from the overall income distribution. Actually finding those numbers seems difficult, unfortunately.

Running with the "private cities as hotels" metaphor quickly turns ugly:

Funny, I have never found much medium quality stuff in the US of A, compared to, say, France or Germany. The US middle class never did have any taste. I don't think there's anything deep going on. The surmised phenomena might simply be the consequences of a high unemployment rate and a drop in labor force participation, and are not inconsistent with what Ironman says.

i wonder if this is in part a preference bifurcation -- buy your staples/ordinary items cheaply, buy a few status items at very high end stores.

I have noticed for years that it is very hard to find "decent" quality goods at the right "decent" price. Normally there is a vast multitude of cheap crap and a sizable segment of boutique expensive goods.
I know that the definition of "decent" is very subjective but in my case it means something that doesn't need to be replaced after a year of use.

And the "cheap crap" isn't what it used to be. My mom bought a lot of my clothes at Sears when I was a kid and it lasted forever. I bought my son a winter coat there a couple of years ago and it didn't make it through the season.

Does the expensive clothing last forever?

I briefly worked at the HQ for a higher end clothing company when I first got out of college in their operations department. One perk was very cheap (or free) acquisitions of their products. That was in 2001. I still have a few of their shirts, sweaters, and pants in my dresser. And due to their simple and "classic" look, they're actually still "in style." But mainly, I'm just happy that 10 years later, I can still fit into them (although the pants are a wee bit tight).

I was actually just thinking about this because their HQ in Manhattan was where I was working on 9/11 and I was kinda surprised to realize that I still wear a few items of clothes that were in my wardrobe on that day as I watched the towers fall from our company's break room.

I have to disagree. I've been finding excellent low cost products over the years. Cheap is no longer always crap. And Chinese goods are not all trinkets. I must say even Walmart has pleasantly surprised me with some pretty nice finds.

This whole phenomena of the "disappearing middle" with regard to products was described in the book "Trading Up" by Michael Silverstein (published in 2003). It's just happening now on a wider scale.

Let's just stick with Obama's definition: if you have a job, you are rich. If not, you are poor.

The best thing about this is that, by definition, there is no middle class. You are then free to endlessly complain that there is no middle class, and that the rich must therefore pay more.

It's really quite brilliant.

One thing that I suspect, based on anecdotal evidence from my wife who sells small businesses, is a bit of "hourglassing" among both rich and poor. Many wealthy people shop at Walmart and Dollar General, and even more at more "affluent-friendly" places like Costco and Target. But they are perfectly happy to buy bulk, cheap stuff if they don't think there's a quality difference. Affluent enviros will spend more on "green" brands, but many others (particularly rich immigrants) will happily buy whatever's cheapest unless it's something "visible" (like cars or clothes) or has a perceived quality difference (like skin-care products).

I agree; it seems that the low end is simply becoming more popular with the middle and upper classes. There is less of a stigma to buying low; with Target and the like people instead seek to brag about their "bargains."
And then every once in a while splurge on the expensive items.
So I see it as less a withdrawl from the middle then following the consumers to the hourglass.

dan1111 said:
As a conservative, I do find it kind of annoying–that people keep asserting this without arguing for it. Here are U.S. census data:
The inflation-adjusted income of every segment of the American population has consistently grown over time.

Since this is household income, it overlooks the effect of the 'Two paycheck' household that has become more common over the last few decades. In 1960, most women didn't work. Now the majority do, at least after the kids are in school. So 'middle class' families have had to double up to maintain incomes.

Then there are the major economic shifts of the last few decades - medical and college expenses have increased faster than the rate of inflation, and private sector pensions have all but disappeared (except for the top ranks of corporations who arguably need them the least) forcing workers to finance their own retirement. Thus 'middle class' households face greater costs to stay in and prepare their children to participate in a 'middle class' lifestyle.

Those with higher incomes have more ability to take these increased costs in stride, and still accumulate wealth. Those in the ‘middle class’ are lucky to have enough left over after putting the kids through the state university to avoid cat food as a retirement staple.

Keeping these changes in mind, look again at the data. The median (50%) income has increased ~30% from 1967 to the peak year of 1999, and has been even lower since then. The 90th percentile income peak was in 2006, and was almost 70% higher than 1967, and the 95th percentile income was 73% higher. I think a good argument could be made that the 30% median gain (and maybe more) was eaten up by increasing medical, education and retirement costs.

Does this qualify as 'arguing' the assertion that the 'middle class' is under economic pressure?

How much of this is just the result of ignoring both ends overmuch? I could see a society in which, because firms thought the money was in the middle, that market had been completely saturated leaving the upper and lower portions largely open.

Comments for this post are closed