Asymmetrical Information

Because most journalists don't have UChicago MBAs and this sort of thing makes for good headlines.

Nice timing...Us wonks at Heritage were looking at the exact same numbers and trying to find out. Placed a call to the DFP (and were asked if were Buck) to try to understand what they did.

We were going to call Census to ask if they could do a special tabulation of the ACS to represent the '99 income figure and maybe that is how the DFP got the '99 income figures.

I was also trying to figure out why they picked '99 as a baseline year and that explains it.

Your sleuthing has made my day much easier so thanks to you and Buck.

It's things like this that impell me to tell my high school and college age sons (both taking statistics courses) to be very wary of how people use data. I was taught "figures lie and liars figure." And while that might be a bit harsh, I think it's still good to keep that in mind...particularly in the "silly season" running up to elections.

I was taught "figures lie and liars figure."

I believe the original quote was actually, "Figures don't like, but liars can figure". Presumably to simultaneously criticize the average person's implicit trust in statistics, and those persons who would misuse numerical data for fraudulent ends.

To paraphrase Homer Simpson:

Oh, Jane, statistics can be used to prove anything. 75% of people know that!

Hmmm, the Freep's Truly Appalling Statistics just happen to fit right in with the left-wing media's preferred story line. Go figure.

I want to hear about this "idiosyncratic creative project". I'm intrigued. I want to take off work and do that. What could it be?

AS: I don't think I'd go that far in implication (that it's a deliberate attempt to further such a story).

More likely the DFP people already had (from reading said media, and besides, as Jane said, median income did drop a bit post-bubble) the idea that a drop was real, and then combined the two sets of stats without a huge amount of care - because they're not economists or statisticians - and didn't double-super-spot-check because the results matched what all their other data told them.

Sloppy and lazy, but not sinister or calculating. That's a more parsimonious, simpler, and at least as consistent explanation, and thus I prefer it. (Never, as the saying should go, attribute to malice or calculation what can be explained by mere incompetence or laziness.)

the point is not that it's sinister, it's that they're more likely to make this sort of sloppy error because they already know the story they want to tell (they think of it as the truth), doesn't really matter which set of figures you use to back it up, because all reasonable/educated people know it's true. called confirmation bias, i believe.

the point is not that it's sinister, it's that they're more likely to make this sort of sloppy error because they already know the story they want to tell (they think of it as the truth), doesn't really matter which set of figures you use to back it up, because all reasonable/educated people know it's true. called confirmation bias, i believe.

That's a weak charge, for two reasons:

1) Pretty much all newspaper and magazine stories come about because someone has pitched a story they think might be worth looking into. No journalist goes about randomly looking at social science datasets in the hope that something interesting might or might not turn up.

2) There is actually a fair amount of debate on the left half of the political spectrum about whether and to what extent there is increased economic inequality and if so why. (In particular, Brad DeLong has been arguing with some people about it.) If anything, the confirmation bias is on the right, where commentators -- with some exceptions -- just don't talk about the subject.

ummmmmmm, isn't that exactly what i'm saying? the story pitched is how bad things are, it fits within the current lefty discussion of inequality, so they find things that support that story line, and likely don't check too hard to make sure they're getting it straight, or don't worry too much about the misimpression created by comparing apples and oranges (because the figures just serve to confirm the point they want to make). don't get me wrong, i suspect some of it's also incompetence/sloppiness, which happens everywhere.

somewhat related: what i find funny is that much of the media is so clearly trying to sell what they think their readership wants -- that's what businesses do -- but ignore that they may be turning off a significant chunk of potential readers who don't buy whichever party line they're pitching.

alkali,

As someone who frequently reads the left side of the blogosphere, I can tell you there isn't debate within the left about whether there is inequality, or even if the inequality is increasing or not, and there isn't much debate about what they want to do about it other than how to raise taxes on the "rich". To the extent they debate the issue, it is with commenters like myself that are outside the left, and half of that debate isn't really debate, but rather consists of personal attacks or suggestions that if I don't like paying higher taxes I can just pack my stuff and move to Somalia. It is only on the right that I see real debate on the subject.

Sooooo.... she's ADMITTING to and act of unconscionable dishonesty, but not that it was wrong?

Yancey Ward: I would refer you to this post at Brad Delong's site, which encapsulates a lot of the discussion (including some with non-lefty Andrew Samwick). This comment from James Galbraith is worth reading too.

Did they count illegal Mexicans? That number will fall off a cliff if they're legalized.

You have the real data. Post it.

Amazing to realize that there are people as stupid (well, uneducated) as this Victoria Turk. But don't expect to see this sort of statistical/social science illiteracy featured in the "Why Can't We Have a Decent Press Corps?" parade at Brad Delong or Crooked Timber; that would violate the first principle of the Popular Front.

DeLong's gimlet-eyed scrutiny of The Map:

There are problems with median earnings numbers, but if it were a great market to be a worker, percentage changes since 2000 would not look like this:

With skeptics like this, who needs dupes?

If you pick the "correct" time interval and statistic, then assume that the economic behavior is due to any recent change in policy and/or party, you can make an argument for almost anything. Measure from the trough of a recession and you always get an increase in gdp, income, etc., measure from the peak and you get the reverse. This is more honest than claiming that Bush's tax cut is responsible for the growth since 2003. At least they took peak to peak (assuming this is near the peak). Since we have had about a 10% increase in productivity over the time period having any fall or even a smaller increase in income shows some distortion in the income distribution. This look to me to be a trend that has been going on for decades. For graph of time series for various income measures see
http://www.visualizingeconomics.com/

dj: Yes, that's what I was saying.

I just thought your words implied more than they should, but obviously, given your response, you didn't mean them to. Happens. Stupid "language".

Alkali: Brad DeLong does not count, by himself, as significant debate, and confirmation bias is not normally related to "not talking about it". That's "not being interested in the subject", mostly.

This is a pretty big stretch to try to inflict doubt in some pretty clear numbers.

Why is it so important to you? Isn't it better to acknowledge that the Bush administration has been bad for the finances of average families and FIGURE OUT HOW REPUBLICANS CAN AVOID MAKING THESE SAME MISTAKES IN THE FUTURE?

Why defend policies that look fine on Heritage Foundation reports, but in reality are gutting the middle class?

We need to know the truth, not spend hours trying tobsfuscate the truth in order to prop up bad policy. We need more honest thinking and less knee jerk defensiveness.

jeez,
I hope you don't feel too bad when I say it took me a little time to figure out whether you were being satirical or not. Did you even understand what she wrote?

P.S. Just in case you were being satirical, my bad.

By the way, jeez- excellent example of confirmation bias.

Let me get this straight: The FREEP had data from the same survey for 1999 and 2005. The FREEP had data from one survey in 2005 and a different survey in 1999. The FREEP choose to use the data from the two different surveys for its study. This decision results in a reported decline in average household income that is an order of magnitude larger than you'd get by using the same survey's data. Did I get that right? Now, the FREEP wants us to believe their decision was not motivated by a desire to make their findings as dramatic as possible. Sorry. No sale. I don't know that they made the decision they did out of political motivations. They simply may have thought this was a better "story" and would sell more papers. Doesn't matter. What was done was wrong and is another example of why what the press says cannot be taken at face value.

Washinton Monthly has posted the data.

I have a post in the to-be-inspected-by-the-proprietor hopper here that contains my own cut at that data. Looking at the Washington Monthly chart, I realized I misread the underlying data by using current rather than real figures. So ignore my post (or perhaps JG will choose, mercifully, to moderate it out of existence).

Census data is not data from a census. It is data from a sample in which the respondents are self-selected. The sample may or may not represent the universe and probably doesn't. For example. census data does not provide accurate on illegal aliens because they lie because they fear arrest. People who do decide to answer the questionnaire do so for selfish reasons and often provide data which is incorrect.

One way to check the accuracy of the data is to compare data from the same households over time. It is impossible to get unweighted census data on a disaggregated basis questionnaire by questionnaire let alone get the hh members identified.

No effort is ever made by the cesus bureau to verify the accuracy of the raw data. Projections from the raw data do have a sound statistical basis but the end product is always "massaged" to compliment current politically correct weltanshuang. Even the most basic numbers, ie. number of citizens broken down by age and sex, are massaged to agree with expectations.

Data collected by the Census Bureau is not good enough for any meaningful purpose but it is good enough for its constitutional purpose which is to assign the 435 votes in the House to the States. Otherwise it is more fiction than fact.

Altho our household income is below the median, and well below the median for our age bracket, we are living well, and living better than we did in 1999. Some of this is clearly due to the Bush tax cuts, because the work I do is on expansion projects financed by capital investment. I don't give a couple of data points in a shoe box about any deliberately massaged larger picture.

Good catch - careful checking of data is an important but rarely rewarding activity.

As a side note, I struggle with "American median income has dropped a little over the last six years, thanks to the popping of the technology bubble and a surge in non-wage compensation costs (read health care)"

I buy that the growth in health care costs as a share of compensation would lower wages. The technology bubble argument seems pretty hard to follow though - particularly since we are talking about _median_ wages and overall growth has been fine.

The obvious candidate that Jane left out is the rapid growth in corporate profits.
http://www.economist.com/opinion/displayStory.cfm?story_id=3645126

Tom

One factor to consider is that between 2000 and 2004, the average household size declined by 2% from 2.62 per household to 2.57 per household. In addition, the number of family households went from 69% in 2000 to 68% in 2004. (Non-family households went from 31% to 32%) What we may be seeing is a shift from two-earner households to single-earner households. As a result, the average household income will be dropping while the per capita income may still be rising.
Source - Table 55 at this link - http://www.census.gov/prod/2005pubs/06statab/pop.pdf

There has also been a population shift, where the number of households headed by people 55 and older grew by 13% between 2000 and 2004, while those headed by people under 55 only grew by 3.7% Consequently, a larger percentage of the households are being headed by retirees, who will have a lower income, though they may have higher assets. (Table 56)

These types of factors would have to be taken into account in determining whether people are getting poorer, or we are looking at demographic shifts.

Joe.. I understand your underlying comment, but on the surface a 2 earner household of 2.62 people is probably better off than a 1 earner household of 2.57 people

statistics can say what you want them to, but honestly it is pretty obvious we are seeing a demographic shift (shift into baby boomers retiring and dying) and an income shift (significant disparities arising and median incomes not growing since 1973 in real dollars) , and I cannot honestly be optimistic about the outlook for my kids and grandkids...

vinton81 - Yes "a 2 earner household of 2.62 people is probably better off than a 1 earner household of 2.57 people" but that is not what we are looking at with these statistics. We are looking at an income change of a couple percent and demographic changes of a couple percent. We aren't talking about cutting the number of wage earners in half while only reducing the household size by 2%.

Frankly, I don't know whether the income would be rising or falling if things like health insurance and the demographic factors are added into the equiation. But unless we do look at these other factors, the household income as an isolated data point doesnt tell us much.

Vinton:

what you're missing is that there is a large, long term demographic shift towards single person households. Not just single income households but single person. Look for articles about the rise of single women buying houses/condos, etc.

You can have a lower median household income with a higher per cap median income by splitting 2 income households into 2 1 income households. This needn't be because of divorce, but rather from kids moving out (mom doesn't work, dad does, and with a better job the kid moved out... )

Datasets this large are really, really low quality. The best sets to use are IRS sets, though there are problems there as well taht can't be solved without some serious breaches of privacy and hard corellation work.

Statistics confirming someone's biases or policy positions/preferences need to incorporate very thorough examinations of convolving factors if they are to be of any value. Something this complicated is an especial problem!

sol, the point is that when you compare different points in time, you have to use the same study. Self-selection is a problem in every study, but with the 1000s of people the selection biases will even out among groups or at least bias the same groups from study to study. If rich people typically respond 10 percent of the time while poor people respond 8 percent of the time, then each survey will have that bias built in.

So say if the census bureau always biases incomes upward by 4000 dollars and the ACS had the incomes right, then comparing the earlier census data with later ACS data would have a built in drop of 4000 dollars that DID NOT OCCUR.

To drive this point home further, if the ACS did have data from 1999, then I could just as easily use ACS data from 1999 with Census data from 2005 and create a 4000 dollar increase in incomes which also didn't occur.

I wonder what happens to the data if you flip the numbers. Right now they have the "good" data, or nice shiny apples at the front, in 1999. They put the "bad" data or the oranges at the back, or in 2005. Then they make their map from that.

Why don't news agencies ever make the mistake in the opposite direction? How about they put the apples in 2005 and the oranges in 1999 and see how their map looks then? I'm sure the reporters in question would have a big problem with that.

Notice that with the new, correct data posted at the Washington monthly, there are some states that were just HAMMERED. As I count em, 8 states had wage losses of over 10% during this time.

Essentially all of the midwest has lost extremely large amounts of income during a full, extended recovery. Wisconson has lost over 16% inflation adjusted during a period when the economy has grown by over 10 inflation adjusted.

We don't have to wonder why Indiana might actually vote Democrats to office. They are simply voting with their wallets.

In many ways the actual data is a worse picture for the economy than the incorrect data.

Jane,

There are so many its hard to single out one. And its not that any single one is so bad, its that he selectively uses every statistic he can to paint a picture that simply isn't true. A common claim of his is that this economy is the 'Greatest story never told'. Its possible he has said this every time he has a program about the economy for the last 5 months. He says it nearly every time, in any case. He then quotes a few of the current days statistics to back this up, usually quite selectively, for example using whichever employment number was higher over the last month.

Take a quick look at the Washington Monthly numbers. Would you in good faith call this 'The greatest story never told'? But these numbers are exactly what you would expect when you know that median wages are down 2.3% across the country over the last RECOVERY. Its no secret that real incomes are down.

People are suffering out there, during a recovery. There are real people behind those statistics. At least the Freep numbers attempted to explain this reality to their readers in Michigan, whose numbers are down 15% btw. Do you think this would be news in Michigan? Hell yeah. For the people of Michigan or even the marjority of people in the US, is this 'The greatest story never told'?

Even though they were incorrect in their methodology, they were trying to explain reality as experienced by a majority of their readers.

Larry, well he just lies to further some strange agenda.

This whole discussion is meaningless anyways. People attach far too much importance to the President over the business cycle, when the credit (or blame) should really go to Greenspan.

Great debunking. Thanks- I too saw the graphs and as a financial analyst I knew they were wrong. The media has stopped to a new low here in Philly. They headline every liberal think tank who makes up stuff about the economy is horrible.

Even though they were incorrect in their methodology, they were trying to explain reality as experienced by a majority of their readers.

Nice try Bill.

I'll put you in the prefers lies over attempts to explain reality camp.

Both astrology and witchcraft were precursors to science. In fact, many of the early astronomical measurements were made by the ancients in the service of astronomy. The herbalism of witchcraft eventually evolved into botany. Today, we have better ways to explain nature.

Larry is more like a modern astrologer who got there by studying astrophysics. He is someone who by their training should know better but refuses to accept reality.

Look at the states whose incomes went down the most; Texas, Arizona, Nebraska, Kansas, New Mexico, Colorado. All of those states have huge populations of illegal immigrants. This is the median (or average) houshold income. Huge numbers of unskilled immigrants have a bad habbit of reducing your average household income. I would like to see the numbers sans immmigrant households. My guess is that the natives did quite well.

Mickslam,
Wisconsin may not be as bad off as you think. Here are two press releases from the past 10 days:

"Data from the Bureau of Economic Analysis (BEA) and Bureau of Labor Statistics (BLS) strongly disputes the Census Bureau data released yesterday. In fact, these widely used national sources for wage and income data show that Wisconsin’s total personal incomes have been demonstrating strong growth. ...
"'The Census Bureau report on median incomes simply isn't a complete picture of what's going on in Wisconsin,' Revenue Secretary Michael L. Morgan said. 'What we do know from sources that are widely used is that wages and incomes have been increasing and that they will continue to do so.'
http://www.dor.state.wi.us/news/060830.html

"MADISON – A growing Wisconsin economy contributed to a 5.6% increase in State General Purpose Revenue (GPR) in Fiscal Year 2006. The Wisconsin Economic Outlook report released today by the Department of Revenue shows employment and incomes in the state increased in the past year. Total personal incomes increased 4.5% in 2005 and employment grew at 1.2%. State GPR totaled $12.03 billion, up 5.6% from Fiscal Year 2005 collections of $11.4 billion. This increase was $80.5 million above the Legislative Fiscal Bureau's January 2006 estimate of $11.9 billion."
http://www.dor.state.wi.us/news/060906.html

The sad part of the whole affair is that nobody will notice that bogus data was used to present a false picture of household income in America.

Is there anyone left who doesn't believe that there was a political agenda behind the way this chart was engineered and is currently being exploited?

There are probably many reasons why household income might be shifting downwards on the median - starting with the fact that millions and millions of very poor Mexican immigrants are being allowed to stream into our country thanks to a tax structure which makes it impossible for people to have large families.

Combine that with the aging of the baby boomers (retirement incomes are almost always lower than pre-retiement incomes) and the fact that median household incomes are declining isn't surprising and doesn't paint the picture the graphic attempts to paint, which was "Bush is evil. Elect Democrats."

Even if true, I don't think there is enough information in median income comparisons to form any valid conclusions. For example: during this period, my three kids grew up and moved out. At the beginning of this period, we were one household with income of (let's pretend) $100k. At the end, my kids, having each gotten jobs paying $50k and my income having doubled to $200k, our median income declined from $100k to $50k. Disaster! Medians are not averages, and household income has an uncertain correlation to per capita income. As someone said earlier, you can't understand this statistic without an understanding of the underlying demographics.

Sorry, I don't follow any of the blogs you guys are referring to, so don't quite understand what the original assertion is -- it's that median household income fell between 1999 and 2005?

If that's the case, the whole thing seems like a non-issue. Median income has fallen since the peak of a completely unsustainable bubble? Well, of course it has!

"Even though they were incorrect in their methodology, they were trying to explain reality as experienced by a majority of their readers."

Pretty sure it's quite the minority that have lost their jobs or suffered pay cuts. "Colorado" may have lost some % of total income over that time but my own has more than tripled.

When I look around I see a very healthy, growing economy. Good statistics, bad statistics, or no statistics be damned, I trust my own eyes.

Even though they were incorrect in their methodology, they were trying to explain reality as experienced by a majority of their readers.

So you're saying the data was fake, but accurate?

Even though they were incorrect in their methodology, they were trying to explain reality as experienced by a majority of their readers.

In other words, the data was fake but accurate.

Oh, come on now- no one posted the best phrase ever "there are lies, damn lies and statistics". At least I did not see anyone mention it in my perusal of the comments. Anyway, everyone have a great weekend. Best regards, Steve Thill

Dang, Timothy beat me to it. That's what I get for reading all the comments without refreshing.

Noah- who are you going to believe, them or your own lying eyes.

Median income is by definition the income where half the people make less then this and half the people make more then this.

If we are adding lots of new jobs, wouldn't one expect that to lower the median income? People who have been in their job or at their company for a long time usually make more then someone just hired for a position.

Good point Ed. And I'm wondering if this could be fairly correlated to the fact that the minimum wage has not been raised for awhile.

There've been tens of millions of immigrants from central america in the
last two decades. The overwhelming majority didn't even go to high
school. Why should we be surprised that median income has been
dropping?

What would be amazing is if it were rising. Neat trick. Difficult to
pull off.

Average income has been rising. Median income has been falling. Falling
median increase is a reflection of the increasing disparity between high
and the low income groups. Rising average income is a reflection of the
increasing prosperity of the country as a whole.

I'm struggling to imagine even a theoretical scenario where tens of millions
of people immigrate to the U.S. and take, and are only qualified for,
relatively low paying jobs and yet somehow we'd see median income rising.
It seems damn unlikely.

As for Victoria Turk and Marisol Bello, I don't think they should be given
the benefit of a doubt. I believe at some level they knew that what they
were doing was deceptive and disingenuous. And if it was a sincere mistake
then we'll see a follow up article from Detroit Free News apologizing to
their readers and explaining what the real situation is. It'll be on the front
page too.

That is, if it was a sincere mistake.

Ed, going from making zero to making minimum wage is an increase in my book. The real question is "are the zeros included?"

Where are the figures for total income and mean income? Or the distribution of income? Does this include the value of entitlements or net income paychecks or gross income?

Too many questions.

The variations of houshold income seem cyclical, driven by the business cycle, but with a long term upward trend.

This is an excellent chart of the situation -

http://www.census.gov/hhes/www/income/incxrace.html

We hit bottom in this current cycle in 2004, 2005 saw a slight increase, and 2006 will probably show strong growth, pretty much on schedule.
Historically speaking, from this point we can expect several years of increasing household income.

"Nice try Bill.

I'll put you in the prefers lies over attempts to explain reality camp. "

Because he doesn't like poorly used inaccurate comparisons; you state that he prefers lies?

So if he liked to massage data and use inaccurate comparisons to shift the results one way or another then he wouldn't be one who "prefers lies"?

Jeez, the irony of that hurts my fillings.

Data such as the census data is so useless, that I basically give it no weight at all. We are talking about surveys in which people self-select themselves to some extent, and the data may or may not actually reflect real incomes since people have more incentive to lie to the government on the low side rather than on the high one. In addition, this data is data for households, and the composition of households is changing over time. In addition to these factors, the demographics of the population is changing, with more retirees spending assets rather than income, for example.

And even if median incomes weren't dropping, the class warriors would find some definition of income that was dropping.

Of course the explanation is bias. If they were interested in making a fair comparison, they would have compared 2005/2000, so they would have a consistent data set, rather than 2005/1999 with its disparate datasets.

Why are you wasting your time on this? A 10% loss in median family's income in ONE year? Tell me it took less than 2 seconds for your brain to process this one.

Good catch -- using inconsistent series without explanation is definitely bad form.

On the right, there is much debate over income inequality; the debate is not whether it exists but whether it matters. I'm on the side that says it does not. Income equality only comes at a hideous and unacceptable societal cost. We will always have rich and poor. The important thing is to give the poor as many opportunities as possible to improve their lot. If there's anything despicable about the rich, it is those who, like Teddy Kennedy and Paris Hilton, inherit it. Those of us most able and most willing to work hard and take risks should reap the rewards. (And no, I'm not rich; in fact I just lost my job. Outsourced.)

It's hogwash from a left-wing press. These are the best of times, by the measures that really count.

Plus, the unemployment rate is lower than it has been for 35 of the last 38 years. Those who insist on pessimism are reducing their own prosperity through risk aversion.

If a Democrat wins in 2008, notice how they will change their tune almost immediately. If there is a recession in 2010, that is still Bush's fault, even if the dot-com crash that began in late 2000 is not Clinton's fault.

This appears to be one of those "well everyone I know is better off but the received wisdom says that isn't true so it must not" urban myths.

I suspect a vigorous rooting out of the real statistics will show that the middle class is doing just fine, income wise. For instance how are household incomes arrived at? Does it include capital gains for instance? 1999 would certainly be measuring from a peak if it does.
More importantly this does not measure actual wealth or asset values merely income. Statistics like these are worse than useless.

One thing that contributes to a decrease in wages is large numbers of low skilled immigrants- both legal and illegal. A mexican worker comes here and makes $5/hr instead of $1/hr in mexico and that lowers the average American wage. Large numbers of low skill immigrants also lower native born American wages thru increased competition.

If there's anything despicable about the rich, it is those who, like Teddy Kennedy and Paris Hilton, inherit it. Those of us most able and most willing to work hard and take risks should reap the rewards. (And no, I'm not rich; in fact I just lost my job. Outsourced.)

I honestly don’t see where all of the animus against Paris Hilton comes from. It seems to me that she (unlike Ted Kennedy, Teresa Heinz Kerry, Mark Dayton, etc.) has never tried to get the government to forcibly take more of someone else’s money to pay for entitlement programs to buy votes for her preferred candidate. So long as she sticks to making reality TV shows and music videos, let her enjoy however much money her parents wish to leave her.

I often skip the statistics and just go on gut feelings. As a semi-retired individual working as a part time bank teller, I have seen a HUGE increase in parents sending money to childern in their 20-30's over the last 8 years. When I started in the 60's it was possible to live on minimum wage. No steak but an apartment by myself and plenty of beans and rice. Today, It takes 3 people making minimum wage to pay the rent on a two bedroom apt. That is NOT progress for the typical american.

let [Paris Hilton] enjoy however much money her parents wish to leave her.

The question is why it's good public policy to prioritize your wish over the wish of others to let wage-earners enjoy however much money their employers choose to pay them. Income and flat taxes take away much of their money, while Republicans are working to remove those taxes from Paris Hilton. Since you prefer to shift the burden from Paris Hilton to wage-earners, you do need to defend your preference.

As for Victoria Turk and Marisol Bello, I don't think they should be given
the benefit of a doubt. I believe at some level they knew that what they
were doing was deceptive and disingenuous.

Sorry, but the Bush Administration has benefited FAR too much from the economic illiteracy of reporters since 2000 for there to be an alternative theory. Occam's Razor.

What? The DFP compared apples to truck tires to try and prove a lie? Who woulda thunk? Why do you folks even read these outlets of intentional misinformation? I have more important things to do, like rearrange my sock drawer.

So, the reporters at the DFP wrote a story and showed themselves to be unable to distinguish apples from oranges. When pressed, they confidently replied that any seeming discrepancies weren't really important anyway because, well gosh, apples and oranges are both fruit, so it really doesn't matter!

I'm with Vulgorilla. C'mon people, are you really SURPRISED that reporters often don't handle facts well? Many reporters aren't competent in the topics they cover and lack the facility with basic statistics and math to build a story around facts. It's easier to write a story based on ideology or emotions anyway. If they have to use numbers, they use them as incantations and to prove that they "get it".

I think we've let ourselves be led on a goose chase. Answering false logic with good logic is the wrong tactic when it's not about logic anyway. Sheesh, there's not much logic in the press but there IS a lot of pomposity. How about answering with ridicule?

Let me see if have all this correct the price of gasoline as doubled in the past 6 years,food housing,health care and clothing have all gone up and the average working persons salary has not kept pace with inflation. But,all we have to do is include the income from the top 10% of americans who have done so well with the help of the flimflam men in washington and the median income as not decreased at all. Is this a great country or what.

Let me see if have all this correct the price of gasoline as doubled in the past 6 years,food housing,health care and clothing have all gone up and the average working persons salary has not kept pace with inflation.

The average working person's salary frequently includes some sort of health-benefits plan, the value of which has increased in tandem with the rising cost of healthcare. There are also other types of benefits, perks, and income sources that are not measured by 'salary'.

So no, you don't have that correct. There is not enough information in 'salary' alone to determine whether your thesis is valid.

I don't know where you live but, here there are 50 mil or so with no health insurance and probability twice that number who are under insured and for those who have health insurance as a perk more often than not any increase in price is coming out of the salary of the person insured. I live on a salary and if it does not keep up with what it cost me to live then there's more than enough information in "salary" alone.

But,all we have to do is include the income from the top 10% of americans who have done so well with the help of the flimflam men in washington and the median income as not decreased at all. Is this a great country or what.

This comment is pretty typical, no? Bob L. doesn't understand what "median income" means, but he feels entitled to spout off about it anyway.

If only the top ten percent of Americans make more money, then the median income doesn't change, because the median income is not same as the average income.

If the median income is $30,000, or whatever it is, that means 50% of people are making less than that and 50% are making more. So if, say, Bill Gate's income goes up by a factor of 10, it doesn't change the median, although it would change the average by a lot.

Mathew,

It is easy to assume that census data has a sample of 100 million but it really doesn't for every question asked in the survey. Non-response rates are extremely high so that when data is reported at the county level it is often based on samples of less than 50. These unrepresentative small samples are then projected to the county population - usually a 1000 times larger - and then the county projected numbers are revised to be politically correct.

In large population counties, the sample is often broken down into subsets to project city totals and suburb totals, sometimes ward totals which are then added up to get PC county totals.

Next the PC county projections are added up on a trial basis to get state totals. If the state totals are politically incorrect, then the county number are revised until a satisfactory PC state total is achieved.

Thus census data will support political goals but cannot be trusted except as a basis for apportioning the House. It can be trusted for this purpose because each political party has a life-or-death stake in this particular count (but only so far as it affects apportionment).

Remember Census data is a snap shot in time. It is a heavily editted picture. Just because a picture is pretty or because it is done by experts does not make it correct.

Picasso's Guernica is a magnificent work of Art but it is not an accurate historical document. It is merely a very expensive work of art. So is the Census.

The 1.1 percent increase in median income in 2005, which was well below the average gain for a recovery year, was driven by a rise in income among elderly households. Median income for non- elderly households (those headed by someone under 65) fell again in 2005, declining by $275, or 0.5 percent. Median income for non-elderly households was $2,000 (or 3.7 percent) lower in 2005 than in 2001.

In addition, the median earnings of both male and female full- time workers declined in 2005. Median earnings for men working full time throughout the year fell for the second straight year, dropping by $774, or 1.8 percent, after adjusting for inflation. The median earnings of full-time year-round female workers fell for the third straight year, declining by $427, or 1.3 percent.

The poor also became poorer. The amount by which the average person who is poor fell below the poverty line ($3,236) in 2005 was the highest on record, as was the share of the poor who fell below half of the poverty line.

Okay, Bob L, but this is what you said:

But,all we have to do is include the income from the top 10% of americans who have done so well with the help of the flimflam men in washington and the median income as not decreased at all. Is this a great country or what.

And that is rank statistical ignorance. I neither know nor care whether median income really has declined or not. But the median income, by definition, cannot increase merely because the top 10% are getting richer, anymore than it could decrease merely by the poorest 10% getting poorer, and if both happen simultaneously then the median income won't change at all.

And that is all that I was responding to. A lot of people don't know what "median" means, and you are clearly included among their number.

And that is rank statistical ignorance. that's really good Gabreil. But,I have some news for you I did not post the last in response to you. It was just to point out the lie. If it makes you feel better you are correct. I let my anger get the best of me and should of taken more time before I posted the 1st.

Bob L -

Even the figures you post are meaningless. Median figures by themselves are meaningless. Let me give you a simple example.

5 people exist, making (2000, 3000, 4000, 5000, 6000) annually. Median 4000. Average 4000.

Each of the 5 people earn 500 more, and 2 new persons start working who make 2500 and 3000. (2500, 2500, 3000, 3500, 4500, 5500, 6500) Median 3500. Average 4000.

Now, the median has dropped 500 while every person who was working the year before is earning 500 more. In addition, the lowest earner year 2 is earning more than the lowest earner in year 1. So is the median dropping good, or bad? It doesn't mean anything until you know what happened to the actual individuals involved, who has dropped out (like the big hit if Ms. 6500 retired) and who has gone from 0 to low positive.

I know that this oversimplifies the situation, and that we didn't have anything like a 40% increase in employment in those 6 years, but hopefully it gets the point across.

Comments are Closed.