Tuesday, January 29, 2013

US income distribution

Where is the middle class in this figure? If we eliminate young and old people, does the peak shift to the $30k range, or are things bleaker than I had thought? [ See links in comments for better figures. ]



Wikipedia: ... Household income in the United States varies substantially with the age of the person who heads the household. Overall, the median household income increased with the age of householder until retirement age when household income started to decline.[25] The highest median household income was found among households headed by working baby-boomers.[25]

Households headed by persons between the ages of 45 and 54 had a median household income of $61,111 and a mean household income of $77,634. The median income per member of household for this particular group was $27,924. The highest median income per member of household was among those between the ages of 54 and 64 with $30,544 [The reason this figure is lower than the next group is because Pensions and Social Security add to income while a portion of older individuals also have work-related income.]).[25]

The group with the second highest median household income, were households headed by persons between the ages 35 and 44 with a median income of $56,785, followed by those in the age group between 55 and 64 with $50,400. Not surprisingly the lowest income group was composed of those households headed by individuals younger than 24, followed by those headed by persons over the age of 75. Overall, households headed by persons above the age of seventy-five had a median household income of $20,467 with the median household income per member of household being $18,645. These figures support the general assumption that median household income as well as the median income per member of household peaked among those households headed by middle aged persons, increasing with the age of the householder and the size of the household until the householder reaches the age of 64.

22 comments:

esmith said...

The peak is an artifact of the linear income scale. If you restrict to 45-64 year old householders and switch to log scale, you'll get a bell curve with the peak near 70-80k. If you dig down deeper by excluding nonfamily households (i.e. single people) and restricting to families where at least one person has a college degree (part of a standard definition of "middle class", cuts the remaining set of families more than in half), you'll probably get into low six digits.

bob sykes said...

Does this include welfare transfers to low income households?

Cornelius said...

Here are some different ways of looking at similar data.

http://politicalcalculations.blogspot.com/2007/06/distributing-income-by-age-group-in-us.html#.UQkqOWfymK0

http://m.townhall.com/columnists/politicalcalculations/2013/01/27/visualizing-the-2012-distribution-of-income-in-the-us-by-age-n1498712

The distribution looks more normal if you focus down on specific age groups.

Bobdisqus said...

I dislike household income as a measure of economic wellbeing as it hides the social disruption of the % of that number from the primary earner. Multiple incomes should not be needed to be middle class. Yes, Steve things are bleaker than you think. Your Murray bubble distortion field is thick. The outsourcing to low wage countries that has been felt so painfully in manufacturing now accelerates into the professional (get a collage education) jobs. (http://infoproc.blogspot.com/2013/01/us-china-software-arbitrage.html )

Do you know a single one of these people:
http://www.whitehousedossier.com/2013/01/15/long-term-unemployment-highest-level-wwii/

Emil Kirkegaard said...

Looks like an F-statistic, yes? Except for the oddity at the high end. It can be fixed by adding more categories of income. http://images.flatworldknowledge.com/shafer/shafer-fig11_011.jpg

esmith said...

Indeed.
"Seeber is just one high-tech worker who's been a victim of outsourcing, in which
jobs like computer programming are transferred overseas to countries such as
India and China. ... is the United States, so long dominant in the field of
computer science and programming, finally losing its edge because of it? Is this
the beginning of a downward spiral, one that will end with rampant unemployment
throughout the country?"

http://www.atlanticfinancial.com/about-atlantic-financial/news/job_oursourcing.htm

tractal said...

But it increases GDP how could it not be the best thing ever for America help me understand.

LX5 said...

The problem with the graph is that many of the people in the $200,000 per year category are actually people middle class people or even poor. How is that possible? A low income person moves into a high crime area buying a house for $50,000. 40 years later the area gentrifies and a house that was worth $50,000 is now sold for $500,000. They retire and sell the house. They get a $250,000 exception so their reported income that year is the capital gain on the house $500,000 - $50,000 minus the $250,000 exemption which is $200,000. They may have such little income that they were forced to move do to increased property taxes. So I believe that graph is misrepresenting the situation perhaps in an attempt to create a consensus that such people are really rich and undeserving to taxing them is justified. That's my conspiracy theory anyway.

gide07 said...

It's the old "doesn't affect me". The trend has nothing to do with "real" people. Wait until Elon Musk clones himself ;) You and your Moscow State betters will be despicable moochers. For more read this this. It's be a man smarter than both of us but at the same time a pathetic loser and murderer.

gide07 said...

"Multiple incomes should not be needed to be middle class."And they didn't used to be. Thank you elite.

esmith said...

What trend? College education was always part of the definition of the middle class. College educated people always earned more than factory workers and burger flippers.

gide07 said...

Sorry. The stagnation of real wages in the US since 1980 and the increase in the GINI since 1950. Wages for the merely college educated (what did they study?) have gone up but not by much according to this<\a>.

esmith said...

First of all, the GINI index has not been increasing since 1950. It hit an all-time low in 68 or 69 and the strong upward trend did not begin till ~1980.

The common story behind the stagnation of real wages and the increase of GINI is the success of the "capitalist revolution". Top income tax brackets were repeatedly lowered, capital gains were given preferential tax treatment, unions were busted, manufacturing was outsourced, and the share of labor began a steady decline (in other words, the capitalist class became able to extract a larger share of value-added from the economy and to stick workers with less).


It is a general rule that 20% of people do 80% of the work. You can fire the remaining 80%, but it will not change the pattern.

LondonYoung said...

Tiny note - this particular data is based on self-reported income. The census bureau believes it reasonably captures wages and salaries, but all other income is known to be severely under-reported. So, employer provided health insurance, food stamps, medicare/medicaid benefits, section 8 housing, free cell phones (below a certain income threshold everyone in the U.S. is entitled to a free cell phone), etc... are intended by the census bureau to be income, but are usually not reported as such by respondents. As long as you keep this in mind, it is not that damaging to use this data as a "snap-shot", but if you get carried away and try to evaluate income distribution changes over many decades with this stuff you are just wasting your time.

One way to get a grip on how much "income" is escaping is to look at the ratio of "IRS reported gross income" to "GDP" - a figure which is steadily falling. More and more citizens - of all income levels - are getting benefits without running them through a 1040 form. At low income levels think food stamps, at middle income levels think "employer provided health care", for home owners think of "home owner rental equivalent income". That last is quite a doosey - and as home ownership grows over the decades it distorts data like this.

I will also point out that the census bureau intends income to include pre-tax wages, but I don't see how the money withheld from a paycheck for taxes should count as income - at least for 90+ percent of what people try to use this data for ...

gide07 said...

"First of all, the GINI index has not been increasing since 1950. It hit
an all-time low in 68 or 69 and the strong upward trend did not begin
till ~1980." You are correct.

gide07 said...

And when health care costs in the US are measured in outcomes rather than $s? And when mortagage payments are subtracted from the rent avoided? And what % of the budget is foodstamps?

Bobdisqus said...

http://www.businessinsider.com/maps-that-explain-the-world-2013-2#change-in-median-household-income-since-2007-gains-are-concentrated-in-the-shale-and-farm-rich-regions-23

Bobdisqus said...

http://research.stlouisfed.org/fred2/series/CIVPART/

Bobdisqus said...

http://www.bls.gov/emp/ep_table_303.htm

Bobdisqus said...

http://demographymatters.blogspot.com/2012/02/trend-breakdowns-in-us-workforce.html

Anthony_A said...

No, it wasn't. As late as the early 80s, skilled blue-collar workers could make as much as the engineers they worked for, especially since the engineers were salaried and the tradesmen could collect overtime. The neighborhood I grew up in was solidly middle-class, and I'd guess about a third of the householders didn't have college degrees, though most had better-paying jobs than assembly-line grunts.

LondonYoung said...

On immigration - I submit for your consideration that your ancestors actually did move from Europe to the U.S., but I suspect that you will not move from the U.S. to Europe?

Blog Archive

Labels