Asymmetrical Information

On the other hand, those of us who are concered about concentrations of power in society, disenfranchisement, insecurity and the creation of an underclass with no real stake in the maintenance of civil society, we're more concerned about *wealth* inequality than income. And that's much worse.

D^2 is correct. If you're willing to be sufficiently reductive in the way you look at the situation, you can massage away almost any inequality. For example, I bet that the daily food consumption of slaves in the antebellum South, measured in calories, was close to or even exceeded that of their owners.

This nation was not founded on the idea that all people would be given a "fair" portion of wealth. If you truly think that, then you'll always find me placing my vote firmly in opposition to yours. If you actually have one.

I think you're going to have a tough time convincing any respectable portion of this country that the government ought to be more involved in redistribution of wealth.

Alkali, your use of non sequitur isn't convincing. Equating the lower quintile with slaves doesn't do much to bolster your argument, either.

David Perron writes:

Alkali, your use of non sequitur isn't convincing. Equating the lower quintile with slaves doesn't do much to bolster your argument, either.

That suggests that you do not know what non sequitur means. "Non sequitur" means "it does not follow," and suggests that I have drawn an logically erroneous or nonsensical inference. I was not drawing an inference, I was buttressing D^2's point by making a provocative analogy. If you think the analogy is inapt or its premises flawed, you are of course free to criticize it.

Your contention that I was "equating" poor Americans with slaves is of course ridiculous. If you want to criticize the analogy, go ahead, but attributing to me statements I did not make does not persuade me and simply pollutes the discourse.

Perhaps Ann Rice isn't the best way to start this point but...

In the beginning of "The Vampire Lestat" (which I couldn't even finish I found it so dull) there is a lovely two page rant by Lestat about how amazing the world looked in the mid-eights as compared to the 1920s. How pretty much everyone was well fed, peasents are not dying in the streets, everyone pretty much has well made, colorful cloths, and lots of them, and pretty much everyone has a satisfactory level of education and how wonderful this is.

I love that passage because it shows two things. The first, how amazing it is when you make sure that everyone has the basic necessities and people are full and healthy and warm. That's a benefit to all man kind.

But part of the reason that is the case is because many of those items which people didn't have before in abundance (with the exception of health care and housing) are cheeper than they used to be (Jane, my little economist friend, will (I believe) back me up on that. Proportionally people are spending less on cloths and food.)

There has always been a disperate break between poor, middle class, wealthy and Bill Gates. Always will be. It is how our society is defined. Our responsibility is to make sure that people who want to attempt to become the Bill Gates' of the world have access to the knowledge, food, shelter, and healthcare that will allow them, if they have the natural ability, to become the next Gates.

So once again I agree all the both posters (thus far) in this string. I am all for rich people. I would like to be one of the one day. But I think as a society we need to make sure that people have access to a system that will allow them with a great deal of hard work to become rich. So that if they are poor and really work hard, they might too have the where-with-all to change the system.

I sit corrected on the non sequitur accusation, and retract accordingly.

A possible third option about Rice is that she writes fiction and not history. I'm not going to form any significant amounts of opinion re social engineering that's based on fiction, and you probably should think twice before you do so. Still, it's a free country.

I'm not saying anything derogatory about charity, mind you. It's government-mandated charity that bothers me.

The working paper is available at:
http://www.econ.jhu.edu/seminars/Spring2003/perri.pdf

It looks like the corrollary to this debt-fueled reduction in consumption inequality
("The evidence presented in this paper leads us to conclude
that credit markets did indeed develop and that this development was the crucial factor for the
divergence between income and consumption inequality in the last 25 years.")
is a high debt burden, which may become crushing in the face of an external shock.
The poor are definitely not out of the woods.

A factor that many studies ignore - as do the arguments for massive income/wealth distribution - is movement among the income and wealth quintiles. I suppose it makes for a dry economics paper but any study that purports be be definitive on the topic of income and wealth 'equality' would have to consider income and wealth mobility. Are people chronically poor, i.e.; in the lowest quintile for their entire lives or, within their families, for generations?

Absent that information it would be hard to argue that anyone is being repressed and/or disenfranchised by a growing disparity of income distribution.

I have never seen the first bit of evidence that any person in our country today has been denied the opportunity to try to become Bill Gates if they want to try. Perhaps one can make the case that some kids are denied opportunity because they were trapped in horrible schools by a govt monopoly that abuses them. But that's about it.

I've read all kinds of heated BS -- usually from the same folks who say that Iraqis were better off under Saddam -- but no evidence.

Opportunity -- it starts and ends there.

I'm amazed that David Perron thinks he's been legitimately corrected. The non-sequitur is in d squared's post, and alkali said that was "correct". But if there is a cause and effect between "wealth inequality" and "the creation of an underclass with no real stake in the maintenance of civil society", it's in the opposite direction from what d squared is implying.

Did the study take into account the changes in arrangements in how people 'earn' in the past two or three decades. Like the decrease in the number of union jobs that pay the same whether you're on the job for 2 minutes or 22 years, and the general trend that pay is more correlated with experience, hence age, than before ? The rise in compensation from stock options and profit sharing. Or that a rising stock market will increase real inequality on its own, obviously, but even if stock ownership were homogeneous across the population, income inequality will rise since income in this case is determined by the timing of sales and purchases, which will not be homogeneous. Ditto for housing. Or an increase in the importation of unskilled, thus low value labor, will increase income inequality as a mathematical certainty. It seems to me that income inequality for the native born could be the result of an increase in the variance, but not the mean, of individual income over a lifetime, and thus not changing consumption alot.

Looks like an opportunity for Miss Crabtree, er Galt, to pick up her chalk and explain the Permanent Income hypothesis.

Actually, Patrick, I accused alkali of indulging in non sequitur, and he properly corrected me. As much as I hate being wrong, on that occasion I clearly was.

David, it's too bad you weren't around in the 16th Century to advise Queen Elizabeth I about those horrible Poor Laws she was responsible for. What was she thinking, trying to help the least able in her society?

Sentence above should read that the the variance but not the mean of individal annual income over a lifetime, not individual income over a lifetime, is increasing.

"On the other hand, those of us who are concered about concentrations of power in society, disenfranchisement, insecurity and the creation of an underclass with no real stake in the maintenance of civil society, we're more concerned about *wealth* inequality than income. And that's much worse."

This is definitely not a problem in the United States. On the contrary, the exact opposite is occurring. The majority of our people are relatively hopeful regarding their future. Socialists like Irving Howe ultimately had to give up the idea of enticing the hoi polloi to revolt against their “masters.” It seems that too many of these so-called members of “an underclass with no real stake in the maintenance of civil society” perceived America as the land of freedom and great opportunities. Also, I am not merely shooting my mouth off. Every major polling organization has found this to be the case. A radical Liberal may claim that these people are delusional---but that is how they perceive the situation!

Dark Avenger, you've induced me to a most alien state, for me - speechlessness.

Sentence above should read that the the variance but not the mean of individal annual income over a lifetime, not individual income over a lifetime, is increasing.

“it's too bad you weren't around in the 16th Century to advise Queen Elizabeth I about those horrible Poor Laws she was responsible for. What was she thinking, trying to help the least able in her society?”

Actually, Great Britain’s efforts to help the poor often backfired---and caused far more harm than good. This is especially true regarding its agricultural policies. Unintentional bad consequences is often the result of “trying to help the least able in her society.”

"Our sample is composed of all households in the CE who are complete income respondents,
with the reference person between the age of 25 and 64 and who report positive income and positive
total consumption expenditure for the interview year (1972-73 samples) or interview quarter (post
1980 samples)."

There's your problem right there; this is a an extremely bad sample. As pointed out in this post of mine, and to some extent in this comment thread over on Delong's, using BLS consumer expenditure data is a sucker's game. It's entirely self-reported; whatever they tell the researchers they spent is what shows up in the survery - the NYT got burned on this, assuming that the flat effective tax rate visible in the data actually reflected the economy as a whole. That they only use full-reporters just distorts it even more.

I'm curious what this type of analysis would result in applied to another data source, but I wouldn't trust the BLS CE data farther than I can throw it.

"Our sample is composed of all households in the CE who are complete income respondents,
with the reference person between the age of 25 and 64 and who report positive income and positive
total consumption expenditure for the interview year (1972-73 samples) or interview quarter (post
1980 samples)."

There's your problem right there; this is a an extremely bad sample. As pointed out in this post of mine, and to some extent in this comment thread over on Delong's, using BLS consumer expenditure data is a sucker's game. It's entirely self-reported; whatever they tell the researchers they spent is what shows up in the survery - the NYT got burned on this, assuming that the flat effective tax rate visible in the data actually reflected the economy as a whole. That they only use full-reporters just distorts it even more.

I'm curious what this type of analysis would result in applied to another data source, but I wouldn't trust the BLS CE data farther than I can throw it.

The much lesser inequality in consumption distribution implies, as the paper says, that income inequality as measured by "snapshots" is exaggerated. If you sell a house or cash in a pension account or stock investment that has been appreciating for 30 years you go to the top of the income distribution -- for one year. If you are in medical school you are at the "impoverished" income level -- with a lifetime of six-figure incomes ahead of you. If you own your own business and have a either a very good or very bad year you go to the top or bottom of the income distribution, for the moment. All exaggerate the snapshot inequality picture, and by more so all the time. E,g. The poorest 5% by income average in the top 15% by wealth, IIRC.

Also, an economy with increasing productivity will always show increasing "snapshot" income inequality even if its members are as equal as termites, because the incomes of senior workers will move up every generation while the incomes of the young don't by nearly as much. So snapshot data will show the richer ever pulling away from the poorer -- even if *within* each generation equality is perfect.

Lifetime income by age cohort is much less unequal than "snapshot" income distribution, and is tracked by consumption because medical students consume with their future incomes in mind, persons with appreciated assets spend knwoing the have them, retirees with poverty-level income spend savings from high income years, etc.

Moreover, more fundamentally, it is *consumption* that counts -- not income. Consumption is the measure of economic welfare. People earn income only to be able to consume.

So those who worry that growing apparent "snapshot" income distribution inequality may signal a welfare distribution problem should be very happy that the consumption numbers -- which are what count for welfare -- say differently. I've been waiting for months for Krugman to hail this good news!

I once had a professor who worked in the BLS. After hearing his stories, one wanted to add national economic statistics to laws and sausages.

"If you're willing to be sufficiently reductive in the way you look at the situation, you can massage away almost any inequality.

The consumption numbers are *inclusive*, not reductive. They include everything -- more than income even, as consumption can exceed income and does so at all the lower income levels.

"It looks like the corrollary to this debt-fueled reduction in consumption inequality ... is a high debt burden"

Not at all. The paper cites the growing credit markets as "allowing households to better smooth their consumption" ... not to take on more debt on net.

E.g, If you are a senior who is "house rich and cash poor" -- paid off home but no money -- you can now realize some of the home's value as income over the rest of your life through a reverse mortgage. You are simply cashing in some of your savings -- but in a way you couldn't 20 years ago.

"On the other hand, those of us who are concered about concentrations of power in society, disenfranchisement, insecurity and the creation of an underclass with no real stake in the maintenance of civil society, we're more concerned about *wealth* inequality than income."

Yes, Gates and Buffett are both Democrats, and the super-rich in the Senate are all Democrats too (Kerry, Kohl, Corzine, Rockefeller, Feinstein). It is troubling to see how the super-rich always concentrate on the same side in politics.

Jason, the BLS may be a bad sample -- but it's a consistent sample. Sample composition is very important when measuring a snapshot, but what this paper is looking at is variation over time. For which the BLS is more than adequate, unless you know of some methodological discontinuities of which I'm unaware.

"I once had a professor who worked in the BLS. After hearing his stories, one wanted to add national economic statistics to laws and sausages."

There are a lot of legitimate problems with the BLS comsumer expenditure survey, and I wouldn't put a lot of faith in it for any particular number. But nobody I know has ever proposed any systemic bias that would make this result invalid.

I.e, The survey might overstate or understate amounts for specific items, but such errors should be consistent from year to year. (E.g., many think BLS numbers exaggerate inflation by about 1% a year -- but they do so every year, so if you are looking for a trend in inflation the effect washes out and the trend you see is real.)

OTOH, if consumption distribution is in fact widening with income distribution, then the same survey questions must systematically underreport consumption at some levels and overreport it at others by an amount that just offsets this, year after year. Lots of people watch the BLS and I don't think anyone has found such a problem.

"Yes, Gates and Buffett are both Democrats, and the super-rich in the Senate are all Democrats too (Kerry, Kohl, Corzine, Rockefeller, Feinstein). It is troubling to see how the super-rich always concentrate on the same side in politics. "

I find it hilarious when Republicans try to put forth the idea that they are for the common man, the working class, the average joe. It's almost as funny as when they claim that Democrats are the true racists because they believe in civil rights and affirmative action to level the playing field.

The 'super-rich' in the Senate might be Democrats because personal wealth is necessary to run for office as a Democrat. As opposed to receiving $2,000 a head from thousands of merely 'very rich' supporters who are paying you back for the regressive tax cuts you gave them or promise to give them.

"Actually, Great Britain’s efforts to help the poor often backfired---and caused far more harm than good. This is especially true regarding its agricultural policies. Unintentional bad consequences is often the result of “trying to help the least able in her society.”"

David, tell me you don't actually beleive that England should return to serfdom, or that we should return to the 1890s, when labor had no protection, orphans and widows were left to beg, and people worked until they died or starved when they could no longer work.

If that is truly the opinion of conservatives these days, rather than the possible eventual unintended outcome of shortsighted and irresponsible Republican policies, then Krugman alarmism is remarkably understated.

"I find it hilarious when Republicans try to put forth the idea that they are for the common man, the working class, the average joe. It's almost as funny as when they claim that Democrats are the true racists because they believe in civil rights and affirmative action to level the playing field."

When you hear this B.S., reach for your wallet. Sort of like "we're simple country folk" in a business context.

Well, since I'm being looked at, I'll see if I can take a look at this paper sometime this weekend.

Offhand, though, I would certainly expect this to be true. After all, there's a limit to how much you can consume, and if your income doubles from $100 million to $200 million, what are you going to do? Buy another couple of jets?

Anyway, I'll take a look and see what they say.

"Jason, the BLS may be a bad sample -- but it's a consistent sample. Sample composition is very important when measuring a snapshot, but what this paper is looking at is variation over time. For which the BLS is more than adequate, unless you know of some methodological discontinuities of which I'm unaware."

The BLS data shows that its reporters pay an effective flat tax. In reality, people do not pay an effective flat tax. The survey obviously doesn't accurately represent the spending on taxes of the population; why would you think it accurately represents the spending of the population as a whole? This leaves out the whole "using full reporters only" issue.

If you want to say "this group of people looked at by the BLS have had virtually no change in consumption inequality," fine, but I wouldn't go talking about the economy as a whole.

Now that you put it that way, Kevin, I can see there's no reason whatever the government shouldn't just step in and confiscate that extra hundred million. Other than it's stealing, that is.

David,

It could hardly be considered stealing, since the extra, unconsumed, income is of no value, according to Jane and others here.

Of no value to whom? I guarantee that it's got value to the guy who's losing it. Of course, if it's got no value to anyone, then it really ought to be left alone, dontcha think?

Maybe you ought to rethink that statement, eh?

Oh, and I'd be very surprised if Jane would characterize that extra hundred million as having no value. Just a hunch.

My somewhat snarky point was that if you're going to say that it's inequality of consumption, and not income, that matters, then incremental unconsumed income is meaningless, so why not confiscate it?

Now you can argue that someone might consume it next year, and that's reasonable, so maybe it should be confiscated at death. Of course if that's "unfair" then you're saying that differences other than differences in consumption do matter.

"David, tell me you don't actually beleive that England should return to serfdom, or that we should return to the 1890s, when labor had no protection, orphans and widows were left to beg, and people worked until they died or starved when they could no longer work."

The exact opposite is true. I also believe that the Progressive Movement in the United States did a lot of good. Our child labor laws and attempts to protect workers in the working place are to be applauded. It’s a question of finding a prudent balance. There is point where the government must not go any further---or it can cause far more harm than good. That point has essentially been reached today! By the way, the poor in the United States do not remind anyone of characters in a Charles Dickens story. You might wish to take a look at “Myths of Rich & Poor” by W. Michael Cox and Richard Alm.

“...and if your income doubles from $100 million to $200 million, what are you going to do? Buy another couple of jets?”

The overwhelming evidence indicates that the very wealthy invest most of this so-called excess capital. Their money is not just sitting in a locked vault. More importantly, the odds are much higher that this money will be far more productively put to use than handing it over to the government. Does any sane person truly trust politicians and bureaucrats to spend money more wisely?

This is for wallster:

http://www.washingtonpost.com/wp-dyn/articles/A37316-2003Jun26.html

> As sometimes happens with Rep. Patrick Kennedy (D-R.I.), he let his
mouth race ahead of his brain Wednesday night at a gathering of Young
Democrats at the Washington nightspot Acropolis. After presidential
candidate Howard Dean spoke, Kennedy delivered an impassioned peroration
against President Bush's tax cut. We hear that Kennedy told the crowd: "I
don't need Bush's tax cut. I have never worked a [bleeping] day in my life."

"I don't need Bush's tax cut. I have never worked a [bleeping] day in my life."

Yes, and the rest of us have to pay the price for Patrick Kennedy's guilt trip. How much social harm must we endure because some wealthy Liberals feel compelled to prove their "compassion?" The Kennedy family has caused enormous damage to the poor. They self righteously perceive the poverty stricken as always victims of exploitation.

David P: I wasn't making any kind of argument about what to with that extra $100 million. I was just pointing out that since there are practical limits to consumption it's unsurprising that consumption inequality would be smaller than income inequality.

David T: yes, that's true. After all, the only two things you can do with money (aside from putting it in a hole in the ground) is to spend it or invest it. So any money that's not spent is, by definition, invested.

And just for kicks, I should point out that when I complain about growing income inequality, that doesn't necessarily mean that I think higher taxes on the rich are the only possible answer. It's true that I'm not sure why conservatives feel that we should continue to cut taxes on the rich so feverishly, but given a choice, I'd rather try to figure out a way to increase the incomes of the working poor and the middle class in the first place. If we could do that, we could leave taxes out of it.

"I'd rather try to figure out a way to increase the incomes of the working poor and the middle class in the first place. If we could do that, we could leave taxes out of it."

Agreed. I strongly suggest that you immediately obtain a copy of Mickey Kaus' superb book, The End of Equality. It is my adamant opinion that the late (and well meaning) John Rawls unwittingly caused enormous destruction. We should not be concetrating on equality--but instead on making the economic pie bigger for everyone.

"On the other hand, those of us who are concered about concentrations of power in society, disenfranchisement, insecurity and the creation of an underclass with no real stake in the maintenance of civil society, we're more concerned about *wealth* inequality than income. And that's much worse."

I find this fascinating. For one thing, I don't think it's true--if you want to see lots of disenfranchised people who feel on the brink of becoming an underclass, you can come and live in Japan.

For another thing, among the lower orders who do feel that power is being unjustly concentrated and that they themselves are disenfranchised, I'd be surprised if most of them didn't name as the people shutting them out (1) "liberals," (2) the Democratic Party, or (3) some collection of lefty interest groups of perceived world-buggering power.

I'm not saying they're right--I agree with Mr. Dreck that each party is for the common people only to the extent that it gets to tell them what prosperity is and then choose which tradeoffs they make in achieving it. But I take dsquared to be talking not about voting rights but about "disenfranchised" as a state of mind, and I don't think that to most people of modest income, it varies directly by their bank balance.

Because, Bernard, if the rich are investing the extra income, then, assuming that they are at least as efficient at generating returns as the government, then the transaction costs of confiscating the extra income will make doing so a net loss to society.

Jason: go read the inequality study the liberal bloggers have been citing. It also shows an effective flat tax; the very rich pay about 28% of their income due to their much greater ability to shift income into preferred forms. If you leave out the bottom quintile and what I believe they call the "pincer" decile of the top quintile (people who make enough money to pay a lot of taxes, but not enough to shift income significantly), and include, as the BLS does, social security taxes, you'll see that taxes are indeed relatively flat and have been for quite some time.

Again, you're first of all not on as sound methodological ground as you seem to think -- the BLS, like most data sets, has issues, but the fact that Brad DeLong doesn't like it is not dispositive of its utter uselessness; and second of all, you still haven't offered anything that indicates why it is longitudinally, as well as latitudinally, problematic. We rely on lots of self-reported data that we know is imperfect, simply because it's the best proxy we have available. It's not enough to say "there are X problems with the data"; you also have to prove that the problems are either fatal (an argument you'll have trouble making until you build a better proxy), or that they are cause underreporting in the variable we're tracking. You've done neither; the survey is what it is, imperfect but necessary, and you've certainly made no suggestion as to how these problems would specifically effect variance over time.

There are a lot of legitimate problems with the BLS comsumer expenditure survey, and I wouldn't put a lot of faith in it for any particular number. But nobody I know has ever proposed any systemic bias that would make this result invalid.

I.e, The survey might overstate or understate amounts for specific items, but such errors should be consistent from year to year. (E.g., many think BLS numbers exaggerate inflation by about 1% a year -- but they do so every year, so if you are looking for a trend in inflation the effect washes out and the trend you see is real.)

Oh this is good. "There are many problems....but I wont list any. In fact, I'll point to another data set (CPI data) that is not related."

Further, the problem with the CPI is NOT the data, but how the data is used, i.e., how the indexes are calculated (modified laspeyres). There is/was alot of talk to moving towards a geometric means which treats price movements symmetrically.

If you are going to criticize the BLS at least get your damn facts straight.

Oh this is good. "There are many problems....but I wont list any"....
~~~

Well, if it makes you feel any better, start with the big number. It underestimates actual total personal consumption by how many trillion dollars? Compared to the national accounts?

The problem with the Consumer Expenditure Survey data is that it comes from a *survey* in which people self-report without even checking their own records -- the CES doesn't audit people, you know -- so *all* the data in it is subject to the sort of reporting errors that people are notorious for making in such circumstances.

Quick, off the top of your head, how much did you spend on automobile transporation last year?

When the total is off by trillions, you know the components are off too. As is easily verified if one compares the component numbers to data from other "harder" sources.

So one really shouldn't rely on any one number in the survey as being the accurate, true number for the amount really consumed for that item. One should go to other harder sources.

You understand this so far, right?

OK, now all that said, there's no reason that I am familiar with for thinking the misreporting in the CES isn't consistent over time. So that if one is looking for a trend it drops out, revealing a real trend. Much like CPI numbers misreport inflation by about +1% -- but do so each year, so that if you are looking for a trend change in CPI the 1% drops out and the trend is real.

This idea disturbs you in some way?

"If you are going to criticize the BLS at least get your damn facts straight."

Well, Steve, I'm sure you know and can tell us by how many trillion dollars the CES understates actual total personal consumption, right?

And the next time you get into a huff in responding to someone else, try saying something that's relevant to what the other person actually said.

"but the fact that Brad DeLong doesn't like it"

Brad didn't have anything to do with it, it just came up in his comments section.

"You've done neither; the survey is what it is, imperfect but necessary, and you've certainly made no suggestion as to how these problems would specifically effect variance over time."

Full income reporters probably tend to be richer, so that set doesn't match the general population. If you assume the accuracy of reporting is linear with income, then as inequality increases full income reporters become ever more likely to be at the upper end of the distribution.

The way taxes are collected significantly changed over the period; the ratio of state/local to federal taxes was all over the place, and reporting of sales taxes/etc. is probably a lot lower than 1040 payments.

The decreasing marginal preference for consumption income as income increases would naturally tend to underestimate "inequality", of course.

"Jason: go read the inequality study the liberal bloggers have been citing. It also shows an effective flat tax"

I thought it was something like the top 30% of dollars pay 33% of the taxes, while the numbers go wild the farther down you go (bottom 10% of dollars pay 1% of taxes). Which isn't flat at all.

Regardless, the BLS data also shows that all income groups pay a total tax burden of 19% or something, which is a silly result; this occurs because they only use self-reported data.

The site's down or something, but this line from the abstract:

"Furthermore income inequality has increased both between and within education groups while consumption inequality has increased between education groups but mildly declined within groups."

doesn't sound equivalent to "inequality is actually flat because what you're seeing is increased income churn," which appears to be what you're getting at.

David Thomson writes: "Does any sane person truly trust politicians and bureaucrats to spend money more wisely?"

Answer: absolutely not. The argument on this forum is very interesting but as a cause for political action it is lacking. I don't believe the Federal government does even a single thing well. The thought that it should be even more fully entwined in the complex task of reshaping society seems to be a kind of folly.

1) wealth and poverty are partly endogenous. Many of the poor are not in the labor force, for example. If you are a big saver, that will move you into higher wealth brackets.
2) equality of opportunity seems more reasonable to me than equality of outcomes. The tragedy of poverty in the US is that so many don't have access to a quality education. One party seems to advocate choice for the rich, but not the poor.
3) on the 1% bias claim, check out the new superlative index of the CPI. In any case, papers that purport to find bias in the CPI rely on other data sources that have issues as well. The BLS, contrary to popular opinion, is not staffed by idiots, and furthermore gets lots of good advice from academics. Most of the issues discussed here were uncovered by BLS research, and are being addressed to the extent possible.
4) the finding related to consumption is not confined to CES data, nor to this study. At the moment, it is becoming a stylized fact: consumption data reveal that income data are misleading.
5) along those lines, check into the fraction of those in poverty who own washing machines, TV's, etc.
6) inequality gets a bad rap. I'd prefer absolute measures.
7) CES data and BEA data do not, it is true, match up in all instances. But BEA data is loaded with imputations. (For example, most of its figures on consumption are calculated by residuals, i.e. the remainder, what they don't allocate to businesses. Any undermeasurement there then automatically overstates consumption.) So it is unclear that BEA data should be considered the gold standard.
8) we can question the motives of all politicians. But whose POLICY PROPOSALS are likely to aid the poor more? Whose are likely to increase poverty (on an absolute scale)?
9) there is a fair amount of mobility, but not as much as we'd like.

Jason, all you're doing is repeating that there are problems. You've still failed to demonstrate how these problems are relevant to longitudinal studies.

Quick, off the top of your head, how much did you spend on automobile transporation last year?

Answering the survey is timed? I have to answer it quickly and can't sit and think about about how much I spend on average per month?

And the next time you get into a huff in responding to someone else, try saying something that's relevant to what the other person actually said.

I was. You stated the CES was off, and to support that claim you pointed to the CPI which is not biased due to data, but due to how the numbers are calculated. Further, it is not clear that an error in the data has to have the same impact as a known positive bias that is the result of how the data is used in calculating an index.

"You stated the CES was off, and to support that claim you pointed to the CPI which is not biased due to data.."

No, Steve, how would pointing to the CPI support any factual claim at all about the CES? ;-)

I stated that the CES was off without saying how -- then, in response to your loud complaint that I should have, I pointed to the *fact* that it underreports total consumption by *trillions* of dollars. OK? That qualifies as "off" don't you agree?

And then, since you hurled out the notion that I should "at least get the damn facts straight", I asked if *you* could tell us by how many trillions of dollars the CES understates consumption? That is a relevant fact, eh?

I take it you can't. Do you even know how to look it up? And since you're so insistent, here's another relevant and simple little question of fact for you -- by how many trillions of dollars does the CES understate total personal income? Helpful hint: You can look them both up at the same place on the web.

"... you pointed to the CPI which is not biased due to data ..."

The source of the error in the CPI is irrelevant to my point regarding it, as everyone else seems to understand, so your comment is irrelevant. What is relevant is whether the error, *whatever* its source, is or is not consistent over time.

It helps to get logic straight too.

But FYI, as you might realize if you gave it a moment's thought, the CPI overstatement does not result from something as *simple* as the question of what kind of index to use when calculating it, or the smart people at the BLS would have handled that long ago.

The CPI problems derive from a number of difficult issues, including matters such as accounting for the effect of the introduction of new, previously nonexistant products on the cost of living -- which presents a challenging data problem as it must be quantified -- similar quantification problems for quality changes, and so on.

If you are so concerned about knowing facts, I'd think you'd at least have read the Boskin report on this for an introduction to them.

the Boskin report had some good points, but is far from the last word on the subject. (too much heat, not enough light.) check out the new CNSTAT report on the CPI; a lot more measured.

"David, tell me you don't actually beleive that England should return to serfdom, or that we should return to the 1890s, when labor had no protection, orphans and widows were left to beg, and people worked until they died or starved when they could no longer work."

The 1890's absolutely sucked by today's standards, not because the policies then in place were horrible, but simply because the policies put into place since then haven't managed to completely halt material progress over the intervening years.

You can't point to the deprivation of a time period more than a hundred years ago and claim that we today wouldn't have experienced even more growth and advancement had we stuck to those policies throughout the past century.

"It's true that I'm not sure why conservatives feel that we should continue to cut taxes on the rich so feverishly, but given a choice, I'd rather try to figure out a way to increase the incomes of the working poor and the middle class in the first place."

Easy. Start deregulating things, especially those used by the working poor and the middle class. If it can work to make airline travel cheaper, it can do the same for houses and other things.

Also, the biggest thing that the poor lack is police protection for their neighborhoods. This is the government's main job, and it's high time that it is done properly.

"You've still failed to demonstrate how these problems are relevant to longitudinal studies."

Why you expect a shift in a biased sample to accurately reflect a shift in the population at large?

Jason -- you're a math major asking me how a data set can provide longitudinal but not latitudinal accuracy? The sample isn't underweighted; the problem is that people don't correctly estimate what they've spent. But if we're trying to track variance, unless there's a reason to think that there is some factor altering the variance of perceptions, or that we're not representing the sectors where variance occurs, then the sample should be a decent proxy for longitudinal study. If people systemically underestimate what they spend, as long as the systemic bias is constant, then it will track actual variance.

Comments are Closed.