# Income, Wealth and Statistics

• News
Staff Emeritus
2019 Award
I've read a lot of posts, and think it might be helpful to point out some facts that I think would help clarify people's arguments.

1. Income is not wealth. Using one as a proxy for the other is like using velocity as a proxy for position. A small disparity in income, acting over time, becomes a much larger disparity in wealth.

2. Demographics matters. Over the course of a lifetime, people's income and wealth changes. Typically, income rises slowly and wealth less slowly over one's career, peaking just before retirement. At that point, income drops substantially and wealth decreases more slowly. Statistics that are not age-corrected can be highly misleading.

Because of this, a plot of an income percentile (or quartile etc.) does not track a given cohort of people. People move into and out of that percentile. This is even more true for wealth than income. Statements like "such and such percentile gained/lost such and such" do not tell you anything at all about what is happening to individuals. This is even more true for wealth than income.

Additionally, there has been an increase in the number of illegal immigrants to the US. Illegal immigrants make up about 7% of the population (according to the Bear Stearns remittances study), up from about 1.5% fifteen years ago. Today this group makes up a large chunk of the bottom decile in both wealth and income; that was much less true in the past. Any study needs to clearly state how this was treated for it to be interpreted.

3. Accounting is important. Perhaps this is best illustrated by example. Bob taught in the public school system and has just retired. He has a $40,000 a year pension. Joe taught at a private school system and has just retired. His 401(K) is returning$40,000 a year to him. Otherwise they have the same assets. Who is richer?

In most studies, Joe would be considered about a million dollars richer than Bob, even though their standards of living are identical. This is solely due to how we usually calculate wealth - we include defined contribution plans and exclude defined benefit plans. As the fraction of people with defined contributions plans instead of defined benefits plans rises, the calculated wealth disparity will rise, even if the standard of living disparity stays the same.

I hope people will take this into account, and will use this to make arguments like scientists, and not like cheerleaders for their favorite team.

Related General Discussion News on Phys.org
3. Accounting is important. Perhaps this is best illustrated by example. Bob taught in the public school system and has just retired. He has a $40,000 a year pension. Joe taught at a private school system and has just retired. His 401(K) is returning$40,000 a year to him. Otherwise they have the same assets. Who is richer?

In most studies, Joe would be considered about a million dollars richer than Bob, even though their standards of living are identical. This is solely due to how we usually calculate wealth - we include defined contribution plans and exclude defined benefit plans. As the fraction of people with defined contributions plans instead of defined benefits plans rises, the calculated wealth disparity will rise, even if the standard of living disparity stays the same.

I hope people will take this into account, and will use this to make arguments like scientists, and not like cheerleaders for their favorite team.
I'm glad you posted this illustration. Please consider the person who works for minimum wage and progresses to an annual income of $40,000 - but will rely on SS for retirement. Next consider the person who works just a few hours per week, has a lifelong housing subsidy under Section 8, along with food stamps, Medicaid, EITC, Make Work Pay, other child credits (possibly SSDI) - then retires on Social Security (albeit based on the min income contributions) but retains other benefits. If paid from private investments, it would take a tremendous amount of assets to provide the benefits received by the welfare person over a lifetime. The cradle to grave welfare person is comparable to a lottery winner. The person that works at minimum wage from the ground up and can't afford to save - retires on funds contributed through payroll deductions over time. Given the drain on the SS system due to expansion and an aging population - those funds may not be available at the levels the worker had hoped. I say the welfare recipient is richer than the working person - and there's greater job security. russ_watters Mentor Those are issues with the stats themselves. Of equal concern to me is the common implication that wealth inequality is a measure of/proxy for poverty. This comes largely from the fallacy that wealth is a zero sum game: "the rich get richer while the poor get poorer." I've read a lot of posts, and think it might be helpful to point out some facts that I think would help clarify people's arguments. 1. Income is not wealth. Using one as a proxy for the other is like using velocity as a proxy for position. A small disparity in income, acting over time, becomes a much larger disparity in wealth. True but surely some people in that 99% have the resources to save. Consequently wealth disparity would still give some measure of inequality. Of course both income and wealth statistics should be use to get a complete picture. 2. Demographics matters. Over the course of a lifetime, people's income and wealth changes. Typically, income rises slowly and wealth less slowly over one's career, peaking just before retirement. At that point, income drops substantially and wealth decreases more slowly. Statistics that are not age-corrected can be highly misleading. I agree. Perhaps the statistics should be divided up into age groups of 5-10 years. For income, this could be 20-30,30-35,35-40,40-50,50-60,60+. For wealth the 20-30 group is largely irrelevant. Still, it is convenient to be able to communicate the overall picture with just a few numbers. We could average the statics for each age bin and then this should give a number which should be independent of age demographics. Because of this, a plot of an income percentile (or quartile etc.) does not track a given cohort of people. People move into and out of that percentile. This is even more true for wealth than income. Statements like "such and such percentile gained/lost such and such" do not tell you anything at all about what is happening to individuals. This is even more true for wealth than income. Please explain more. Additionally, there has been an increase in the number of illegal immigrants to the US. Illegal immigrants make up about 7% of the population (according to the Bear Stearns remittances study), up from about 1.5% fifteen years ago. Today this group makes up a large chunk of the bottom decile in both wealth and income; that was much less true in the past. Any study needs to clearly state how this was treated for it to be interpreted. So, If we remove the bottom decile from the statistics does this significantly change the results? 3. Accounting is important. Perhaps this is best illustrated by example. Bob taught in the public school system and has just retired. He has a$40,000 a year pension. Joe taught at a private school system and has just retired. His 401(K) is returning $40,000 a year to him. Otherwise they have the same assets. Who is richer? I agree. Of course we can't count all social services as income someone receives. Someone young benefits little from Medicare. Someone on welfare might not consider it much benefit to have to see a social worker or do a drug test to get their welfare dollars. Housing certainly should be counted as income. If Medicare is counted as income it should be age adjusted. Social security is paid for via pay roll taxes, so only what is in excess of these fees should be counted as additional income. I'm curious what studies you think do a good job adjusting for these accounting concerns and how different the results are. I hope people will take this into account, and will use this to make arguments like scientists, and not like cheerleaders for their favorite team. If you want people here to do this then you should show them where they could find the raw data and evaluate their code to analyze the data. My understanding is that these forms do not let the posters do their own interpretation/analyses of the raw data. mheslep Gold Member 2. Demographics matters. ... People move into and out of that percentile. ... We commonly see PF posts stating the opposite is mostly the case, that except for the higher income groups most people are stuck where they start. Long term data demonstrates instead what you state here, individual incomes are mobile in all groups, especially at the bottom. Caveat: for those that are unemployed, and at the moment in the US there are many, the usual methods for increasing income are blocked. Last edited: Staff Emeritus Science Advisor Education Advisor 2019 Award In 2006, these are the figures: Age Median Income 15-24$31K
25-34 $50K 35-44$60K
45-54 $65K 55-64$55K
65+ $28K Age Median Net Worth 20-29$8K
30-39 $44K 40-49$118K
50-59 $182K 60-69$209K

We commonly see PF posts stating the opposite is mostly the case, that except for the higher income groups most people are stuck where they start. Long term data demonstrates instead what you state here, individual incomes are mobile in all groups, especially at the bottom. Caveat: for those that are unemployed, and at the moment in the US there are many, the usual methods for increasing income are blocked.
I think people are saying their is less social mobility. For instance while a University Graduate may get a higher salary they will have to pay off their student debt, wait until their are older to start earning money and will be later owning a house.

Now as for trying to be scientific about it, what statistics do you think best measure social mobility.

In 2006, these are the figures:

Age Median Income
15-24 $31K 25-34$50K
35-44 $60K 45-54$65K
55-64 $55K 65+$28K

Age Median Net Worth
20-29 $8K 30-39$44K
40-49 $118K 50-59$182K
60-69 $209K Do you have a source that gives this over several years? How would you suggest adjusting this for inflation. Shouldn't each demographic see a different rate of inflation? mege Do you have a source that gives this over several years? How would you suggest adjusting this for inflation. Shouldn't each demographic see a different rate of inflation? Why would each group see a different rate of inflation? I would like to think that$.79 buys a candy bar now (or $.59 15 years ago) if I'm 29 or 59. Age shouldn't matter. chiro Science Advisor I've read a lot of posts, and think it might be helpful to point out some facts that I think would help clarify people's arguments. 1. Income is not wealth. Using one as a proxy for the other is like using velocity as a proxy for position. A small disparity in income, acting over time, becomes a much larger disparity in wealth. Absolutely agree, but in saying this what is your definition of wealth? To me a definition of wealth means real ownership of resources. It could be a house, a factory, a business, an IP portfolio or some other resource (even land for example). Of course one would have to clearly define the kinds of resources explicitly, but I think a rule of thumb would be assets that generate some kind of income or something similar. 3. Accounting is important. Perhaps this is best illustrated by example. Bob taught in the public school system and has just retired. He has a$40,000 a year pension. Joe taught at a private school system and has just retired. His 401(K) is returning $40,000 a year to him. Otherwise they have the same assets. Who is richer? In most studies, Joe would be considered about a million dollars richer than Bob, even though their standards of living are identical. This is solely due to how we usually calculate wealth - we include defined contribution plans and exclude defined benefit plans. As the fraction of people with defined contributions plans instead of defined benefits plans rises, the calculated wealth disparity will rise, even if the standard of living disparity stays the same. I hope people will take this into account, and will use this to make arguments like scientists, and not like cheerleaders for their favorite team. Wealth does have a correlation to living standards but they aren't completely correlated. Also value is a very weird thing. It depends on who is doing the valuation since one group could overvalue (making some people happy) and some could undervalue. The idea of some equilbirium being met in a supply-demand free market situation can sound nice, but this isn't always the situation (i.e. there can be interference in the market that causes things to be undervalued or overvalued), so you need to be careful about using some valuation statistics. Again with the definition of wealth above, if you wanted to talk about standards of living you would need to incorporate purchasing power in addition to objects of wealth (i.e. resources that are owned and generate real income). But yeah I do agree with your sentiments about the arguments, but the whole point of these arguments (in my view anyway) should be for everyone to learn a thing or two about something that they don't really know anyway, and even people that do try to be scientific about it as much as they can, still have a thing to learn from people who may not consider themselves scientific, but none-the-less have some solid experience and good points to bring to the table. Not everything is a controlled experiment especially in something like economics and quite frankly I want to hear people with a lot of different backgrounds whether they are business people, employees, or otherwise because these people have a lot to bring to the table despite not considering themselves as a "scientist". Those are issues with the stats themselves. Of equal concern to me is the common implication that wealth inequality is a measure of/proxy for poverty." The attached international statistics are from a NYT article. #### Attachments • 56 KB Views: 454 russ_watters Mentor How would you suggest adjusting this for inflation. Shouldn't each demographic see a different rate of inflation? You wouldn't adjust it for inflation: it is a single snapshot in time. russ_watters Mentor The attached international statistics are from a NYT article. Ok....what is your reason for posting them?  Meh.... I'm going to go out on a limb and assume you posted that to show a positive corellation between income inequality and poverty, contradicting my post. I'm going to go further out on that limb and assume that you weren't aware that the OECD measures poverty using income inequality. So the data you posted is an example of, not a rebuttal of, the problem I was describing. The thing is, it has been my perception that gun to their heads, most people will acknowledge an acceptance of the dictionary definition of "poverty", which is a lack of basic needs. From the OECD website: "Poverty line: An income level that is considered minimally sufficient to sustain a family in terms of food, housing, clothing, medical needs, and so on." But (again, in my perception) this definition conflicts with the ideological desire for fairness through equality of outcome that we've discussed in multiple threads on PF lately and is common in Western society today: Because wealth and income are not zero sum games, you can have both high income inequality and low poverty (or vice versa). This causes some people to stray from the definition of "poverty" that nearly everyone accepts and instead utilize a measure of poverty that attempts to anchor the definition to their ideology. I've seen a lot from the OECD that bothers me, where it appears the organization allows an ideology to influence them to try to force the data to connect their favored ideology to their stated goal. Last edited: AlephZero Science Advisor Homework Helper Why would each group see a different rate of inflation? I would like to think that$.79 buys a candy bar now (or $.59 15 years ago) if I'm 29 or 59. Age shouldn't matter. Sure, the candy bars cost the same for everybody, but your total purchases over a year will show a very different pattern for a 30 year old and a 60 year old. Obvious exaamples: Many 30 years olds will be spending a lot of money on raising their kids. Most 60 year olds will not (except perhaps buying the occasional candy bar for their grandkids). A 30 year old commuting long distances to work will spend much more on fuel and transport than 70 year old retirees. Etc, etc... You wouldn't adjust it for inflation: it is a single snapshot in time. You missed my point. I was saying that if we are going to break down wealth inequality per demographic then we should break down inflation per demographic so the inflation rate used for each demographic represents on average the inflation rate they actually see. I discussed the fact that not everyone sees the same rate of inflation in a previous thread: https://www.physicsforums.com/showthread.php?t=544676 Older people will already own a house so to them housing inflation is not as much of an issue as it is to younger people. Poor people spend a much greater percentage of their income on food and housing so changes in housing prices will have a much large impact on lower income earners then higher income earners. Ok....what is your reason for posting them? "Poverty line: An income level that is considered minimally sufficient to sustain a family in terms of food, housing, clothing, medical needs, and so on." Most people will agree on this but the question is should it be sufficient income to sustain a family or a single individual. If people are dependent on others to live, individuals will have considerably less freedom, dignity and there will be much greater rates of abuse. If only enough income is provided to sustain a family then what size of family should we try to make sustainable? To what extent should the government subsidize peoples choices to have families and how does this affect people that wait until they can afford to have a family. As for food, should they be happy with a high carb/fat diet, or should they be allowed meet which is better for you like say grass fed beef and be able to afford vegetables high in nutrients that is pesticide free? To what extent should they be able to buy pre packaged meals or should they have to make everything from scratch? D H Staff Emeritus Science Advisor You wouldn't adjust it for inflation: it is a single snapshot in time. Sure you could/would. That inflation is a part of the wealth picture (inflation is a cancer that eats into the wealth picture) is one of the justifications for taxing capital gains at a lesser rate than ordinary income. I'm going to go further out on that limb and assume that you weren't aware that the OECD measures poverty using income inequality. The OECD measures poverty using INCOME, not income inequality. Its logically possible to have a high-income gini coefficient society with everyone is above a poverty line. The thing is, it has been my perception that gun to their heads, most people will acknowledge an acceptance of the dictionary definition of "poverty", which is a lack of basic needs. From the OECD website: "Poverty line: An income level that is considered minimally sufficient to sustain a family in terms of food, housing, clothing, medical needs, and so on." I'm not sure the OECD definition is very different from your definition. Certainly I consider food, shelter, clothing and medical 'basic needs'? Obviously, using an income level is a proxy for counting those families who cannot afford food,housing,clothing,etc, but I don't see a cheap way to collect that sort of data without using income as a proxy. But (again, in my perception) this definition conflicts with the ideological desire for fairness through equality of outcome.... This statement is the real reason I've responded to this post. I think the phrase "equality of outcome" (and its bastard brother "equality of opportunity") is a straw-man that no one thinks is desirable. I have no real intuition for this either way- but do the more conservative members of this forum TRULY believe that liberals/progressives/more leftish people truly want everyone to have the same outcome? Because if this is the case, we have serious and fundamental miscommunications. russ_watters Mentor Sure you could/would. That inflation is a part of the wealth picture (inflation is a cancer that eats into the wealth picture) is one of the justifications for taxing capital gains at a lesser rate than ordinary income. I'm aware that inflation is a problem in general, but I'm still not even seeing where one would put an inflation adjustment in a single-point-in-time snapshot. The income data, as it was posted, says something very important to me: Age Median Income 15-24$31K
25-34 $50K 35-44$60K
45-54 $65K 55-64$55K
65+ $28K This says that if I were a typical 22-year old, and the US economy held status quo (flat unemployment rate, just enough growth to counter population growth, etc.) I should expect to see annual 5% over inflation raises for the next 10 years or so, which will increase my annual before tax earnings by a net 61%. Could you tell me, specifically, where you would apply an inflation adjustment and what it might tell us? russ_watters Mentor You missed my point. I was saying that if we are going to break down wealth inequality per demographic then we should break down inflation per demographic so the inflation rate used for each demographic represents on average the inflation rate they actually see. I'm just not clear on why that matters or how you could apply that to the numbers given. If the first bracket makes$31k now and the second bracket makes $50k now, it should not be seen to imply that in 10 years the second bracket (those formerly in the first bracket) will still be making$50k.
I discussed the fact that not everyone sees the same rate of inflation in a previous thread:
I think that's the wrong thread....
Older people will already own a house so to them housing inflation is not as much of an issue as it is to younger people. Poor people spend a much greater percentage of their income on food and housing so changes in housing prices will have a much large impact on lower income earners then higher income earners.
That's fine, but I really don't see how it applies to the age-dependent income data.

I have no real intuition for this either way- but do the more conservative members of this forum TRULY believe that liberals/progressives/more leftish people truly want everyone to have the same outcome? Because if this is the case, we have serious and fundamental miscommunications.
I think the outcome desired by the progressives is permanent political power. Accordingly, they're willing to spend whatever it takes to achieve control.

As an unrepresented independent conservative small businessperson - I favor the checks and balances of a two party system but would like to see the emergence of a third party. I hoped at one point the TEA Party would evolve into a Party of small business owners, managers, and professionals - but the media (IMO) made certain that did not happen.

russ_watters
Mentor
The OECD measures poverty using INCOME, not income inequality. Its logically possible to have a high-income gini coefficient society with everyone is above a poverty line.
Could you source that please? When I noticed the poverty rate provided for the US in Laurie's link was much higher than the US gov't reports, I googled and found this:
wiki said:
The main poverty line used in the OECD and the European Union is based on "economic distance", a level of income set at 60% of the median household income.[21]
http://en.wikipedia.org/wiki/Poverty

That may also be some sloppy wording on my part, as that isn't gini per se, but something related --- still, that doesn't seem to match what you said.
I'm not sure the OECD definition is very different from your definition. Certainly I consider food, shelter, clothing and medical 'basic needs'?
You misread: I'm saying their definition doesn't match their measurement. But until we get the measurement clarified, that might be moot...
Obviously, using an income level is a proxy for counting those families who cannot afford food,housing,clothing,etc, but I don't see a cheap way to collect that sort of data without using income as a proxy.
Clearly, but I also don't think it would be all that difficult to use surveys of those actual quality of life items. But then, if we did a survey for lack of food, lack of housing and lack of clothing, we'd get virtually zero poverty in every developed nation and I don't think that that's a desirable result.
This statement is the real reason I've responded to this post. I think the phrase "equality of outcome" (and its bastard brother "equality of opportunity") is a straw-man that no one thinks is desirable.
I strongly disagree. I think that while there is often some obfuscation, there is no other reasonable way to characterize forced redistribution of resources than "equality of outcome".
I have no real intuition for this either way- but do the more conservative members of this forum TRULY believe that liberals/progressives/more leftish people truly want everyone to have the same outcome? Because if this is the case, we have serious and fundamental miscommunications.
Not exactly equal, no, just more equal (with rarely a definition of how equal). If you thought I meant that "equality of outcome" was intended to be complete, that's not what I intended to convey.

Last edited:
mheslep
Gold Member
From the source referenced in the Wiki article on poverty measurement. Russ is correct:

Within the EU poverty is normally measured by using relative income poverty lines. This involves working out average or median equivalised household incomes in a country. A poverty line is then set which is a percentage of that average income. Commonly these poverty lines range from 40-70% of household income.
http://www.poverty.org.uk/summary/eapn.shtml

To my mind such a measurement would be better defined as the "social engineering line", or the "envy line", with only a vague correlation with poverty.

Staff Emeritus
2019 Award
Now as for trying to be scientific about it, what statistics do you think best measure social mobility.
I don't know the right answer; just that an uncorrected income or wealth percentile is not it.

The attached international statistics are from a NYT article.
Interestingly, they use dollars spent as a proxy for educational quality. Since we have statistics for educational outcomes, wouldn't that have made more sense? (Not that the US does particularly well there either, but at least that directly measures what they want to know)

mheslep
Gold Member
The correlation is worse yet: the NY Times table uses percent of GDP dollars spent.

Staff Emeritus