This article surprised me quite a bit
tl;dr - the UMC has increased from 12% to 30%, defined as $100,000 to $350,000 a year, at least double the U.S. median household income and about five times the poverty level.
http://marginalrevolution.com/marginalrevolution/2016/06/the-middle-class-is-shrinking-because-many-people-are-getting-richer.html
Here's a longer article from the WSJ:
http://blogs.wsj.com/economics/2016/06/21/not-just-the-1-the-upper-middle-class-is-larger-and-richer-than-ever/
As usual, the comments on such things are a garbage fire. Comments here will be watched, discussed, and possibly deleted on my whim. Not for disagreements, but for jerkery.
http://marginalrevolution.com/marginalrevolution/2016/06/the-middle-class-is-shrinking-because-many-people-are-getting-richer.html
Subscribe to:
Post Comments (Atom)
Listening
ReplyDeleteThis seems like a graph designed to obscure more than it reveals. It's adjusted for inflation, but it tracks by household instead of by person. In other words, the expansion of upper middle class income (which I believe they define as $100k+) may actually be the result of families switching to two income households.
ReplyDeleteMark Diaz Truman I'm not sure how tracking by household is obscuring. There's a lot of expenses that are dependent on household size (rent, food), and the only immediate one I see for number of earners is commuting cost. This isn't meaning to say that individuals are getting richer, but that households are....
ReplyDeleteWilliam Nichols It could be obscuring, but only if the poorer generally have larger households now then before and/or the richer have smaller households then before. (Or vice versa, if it's obscuring the unexpected trend of less poor and more rich persons.)
ReplyDeleteOur household technically tips into upper middle class range, because we took on a roommate.
ReplyDeleteLet's say we compare two households: one from 1955 and one from 1995.
ReplyDelete1950: Man works at a factory and earns $10k (roughly $90k today) a year. He works roughly 50 hours a week, and gets a few weeks of vacation each year. His union protects his job, such that he's got virtually lifetime employment (assuming he doesn't screw anything up). His wife attends to all of the social and familial functions of the house. She's taking care of the kids, managing the household, etc.
1995: Man and his wife both work at office jobs. They each earn $40k (roughly $120k today total) for around 60 hours a week of work. They have marginal vacation time, which their employer discourages them from using. Neither of them has a union, and neither of them really has time to maintain a household.
Which of these situations is a better economic situation? According to the metrics in the above chart, our couple is now "in the upper middle class." But I think all of us would agree that the family in 1955 actually has it better based on the socio-economic realities of their position in the economy.
(Quick side note: I'm clearly setting aside some of the gender politics here. It would be easy to imagine a world of single-worker households that aren't gender-segregated. Tough to implement, but the 1955 system doesn't require sexism to function.)
Kelley Vanda As usual, you are as succinct as your are poignant.
ReplyDeleteMark Diaz Truman I think you wanted the couple making 60k each?
ReplyDeleteWilliam Nichols - no, he's letting us know what $80k in 1995 looks like in 2016 dollars.
ReplyDeleteNo, it's 1995. Their collective household income is $80k (roughly 120-130k in today's dollars).
ReplyDeleteAnd yeah... Kelley Vanda nails it. :D
Also consider the methodology used in the study here. A single person household making 58k is considered UMC by equivalency as is a 2 person household making 80k.
ReplyDelete(See Table 1)
Misha B So long as it is double the U.S. median household income and about five times the poverty level, yeah, that'd be UMC by the methodology here. Is there a problem with that methodology?
ReplyDeleteMark Diaz Truman I see. Why are you comparing 1950 to 1995? The chart is specifically discussing societal changes since 1979, so I don't understand the comparison of 1950 to 1995. Why not compare 1950 to now?
ReplyDeleteI'm not sure, but I'd imagine it's because 1950 is often used as the era of the middle class holy grail.
ReplyDeleteSure, where a factory job could get you ahead. I've heard that myth, too. What I don't see is the hard numbers; I'd like to see this analysis carried back to the 50s.
ReplyDeleteBut, what really made me scratch my head wasn't the comparison to the 50s, but the comparison to the 95, instead of now. I'm don't understand the point there.
William Nichols - I don't think the 50s were magical or uniquely special, but I do think that our idea of a strong middle class is warped by hidden costs like the labor women put into managing households or the role that unions had in managing the economy. As my labor economics teacher used to say, "All models are wrong; some models are useful."
ReplyDeleteI picked 1995 because I thought it was important to show how that 40-year gap would result in numbers that sounded reasonable via the study but actually masked a lot of important changes that we're happening in the workforce during the crucial decades in which women entered the workforce (1955-1995). If you want to look at 2016, give both people $60k a year salaries and downgrade their work from office jobs to contract labor. They have to pay for their healthcare out of that amount too. :D
Today, we see dual incomes as completely normal, even desirable, but I think we often overlook how a dual income might have hidden costs. That's not to say that women shouldn't work or that we want folks to stay home. It's instead to note that saying "There are so many richer Americans" might miss how we are all less well off for the hours we work, the job security we don't have, and the new costs we now have to bear (child care, health care, etc). Inflation, in my experience, doesn't account for those things.
Mark Diaz Truman I was about to do a "If I actually read with charity, let's make this 2016, 120k total in income for the couple ... ", so that the comparison could readily be made. Thanks for clarifying!
ReplyDeleteI don't propose -- by any stretch! -- that a chart like this paints the entire picture. But, it does make me question some assumptions, and make me want more data. It is so far afield from my expectations, and from the standard political line -- from either the right or the left (yet published by both marginal revolution and the WSJ) -- that it means I want to reevaluate my understanding. And that's really the value; to question the narrative.
The other thing here is the average family size is falling. One person working out of the home to support six who do not bring in income (wife and five kids) is a very different situation from two working outside the home to support one who does not (one kid). This indexes to family size, but, you're right, the economic cost of child raising is much higher is there is not an adult -- or near adult -- who can take care of them. Cause and effect of this is way too hard too tease out, but kids used to be economically useful by five, and now largely are not. That is a good thing, but may act as a brake on having kids -- people respond to incentives.
I've got a few economist friends -- one working on a similar line -- who I'll talk to in depth when next I have a chance. I want to see the data for myself, but probably don't have the time to do an in depth analysis.
Lastly, I'll take it as a given that you're meaning not to be gendered, and hope you do the same. That is: neither of us would want a situation where only men work for pay and all women are forced out of the labor pool. If anyone did show up with such an opinion, I don't know how long I could be polite. :-)
William Nichols - large families, outside of farms, were never about labor. Religion, infant mortality rates, or a lack of reliable be rth control methods maybe. But not monetary gain. A young child's income might help offset the cost of raising them, but certainly never led to profit.
ReplyDeleteWilliam Nichols - Yes! I say that I don't like this chart because it seems to me to try to answer those questions that you're raising (which are all really, really good) with an overly simplistic. It is far afield of my understanding of the problem (which certainly biases me against it), but I don't think it's far afield in an interesting way. It seems to say "If we cut the data this way, we get an unexpected result."
ReplyDeleteIn economics (and other social sciences), we are especially skeptical of findings like this because they seem like tortured data. Like linking ice cream sales and murders. (True, they are linked... by heat waves.) Something has to match some of what we know already AND expand our understanding of the problem. But sometimes that healthy bias ends up being wrong and we miss stuff, so... it's hard to be tough but fair.
And yeah, the difficulty of discussing economic history is that we don't want to lose sight of the terrible sexism/racism that was baked into certain systems. We've got to find a way to capture what's great about the 50s for white men for everyone. :D
Kelley Vanda Valid. My own folks weren't from the farm, but my grandparents sure were, and at least mine consistently had a farmer mentality -- my dad's parents went back to the farm after retirement. I don't know if they thought of their kids as useful, but I do know that my grandmother had kids until she decided she was done. That's all she said about that.
ReplyDeleteMark Diaz Truman
ReplyDeleteI was similarly skeptical when I first heard of this -- from an economist. That data and paper isn't public yet, so I don't want to go into depth -- but, this is effectively the second independent research with this outcome that I've seen.
I mentioned this to my wife, Dianne Harris , who immediately pointed out that in the 60s, those families had two working adults, but we were only paying one. Now, we (more than we did) pay for home labor, and it no longer supports someone else working outside of the home.
The model provided then isn't one person working 50 hours outside of the home and one person not working -- it is, instead, one person working 50 hours outside of the home, and one person working ~108 hours within the home.
So, I wrote a whole big bunch, but this occurred to me, and it's more important than the whole rest of the garbage I wrote.
ReplyDeleteI've participated in the US Census a few times, and they've never asked me to prove my income. Not once, in all the times I did it.
It seems like one of the least scientific articles on economics I've read in quite a while, to be honest. It's totally based on the honor system.
Why would someone lie about their income on the census? Maybe they aren't lying. Maybe they are just wrong, ball-parking it based on what someone else in the family told them, or what they were lead to believe. Maybe, like Trump, their net worth fluctuates based on how they feel about themselves.
I have to call BS, sorry.
I also looked into the author, and while being a Libetarian thinker doesn't immediately make me think he's full of it, it does make me think he focuses on things that are almost irrelevant to me, and I say that as someone who knows, and loves some RL Libertarians.
https://en.wikipedia.org/wiki/Alex_Tabarrok
Aside from that, something else I was thinking is addressed, shortsightedly, in the comments. So, a few of them assume that this data is authetic, which, fine, but also that the boost comes from people ascending into the investing class.
That's a fair hypothesis, since real work is hard to find these days (and since I'm firmly in the vanishing middle class, I have to volunteer that nobody I know is escaping upwards, it's all been down, and some of my UMC friends have slid right past me on that slope).
The thing is,
1) How many people can just sit back investing, and how does a person reliably break into such an occupation (or lack of occupation)?
Really, it's like being glad that people no longer have to work, because they are doing well in the casinos. That will only last until it doesn't.
2) The most profitable companies do things that are detrimental to most human beings, so having a lot of people invest in them, and no political controls, or otherwise, over their unethical behaviors, is not a net good to me.
Money, seriously, is not everything, and it's especially true if that money is coming from unsustainable practices.
Focus though, on the very first point I bring up, because I think it's important to take this all with a grain of salt, given the amount of rigor was applied to collecting, and vetting the data, which I mean to suggest was exactly none.
They ask for your income the same way they ask for your ethnicity, and in both cases, you can put whatever you want. Nobody, in either case, is going to call you on it, or apply any kind of test.
That's very significant.
T to the E to the O Your hypothesis is that US Census data is crap?
ReplyDeleteThat's interesting, and totally possible. Would you like to suggest a different data source? Failing that, would you like to provide evidence that the US Census data is crap?
My initial research doesn't show anything about that, but is limited to a quick google search. The first result is the census itself, which says its data is awesome. (of course it does). The second is wikipedia, which points out that 2% of whites went uncounted in 1970, and 6% of blacks. At first blush, that seems pretty good!
You're putting words in my... fingers, but let's roll with that, and say I did say it's crap. All of it. I'm not saying that, but let's just say I did.
ReplyDeleteHow would anyone prove that it's not, with regard to data about income?
The Census data is anonymous, and not vetted at all. It's very informal.
Us, "How do you know they made that much?"
Them, "Because they told us so."
Anecdotal Alert!
So, one time, my exGF and I forget to mail in our census, and they sent someone to the door. That person asked what ethnicity we were. We both said that we were Black, White, and Native, because that's the story we've been told by either of our families.
That person looked at both of us, and said,
"I'm marking White for you, and Black for you, because that's what you look like to me."
So, that's the sort of vetting the Census Bureau is capable of. I'm not saying all of their workers are so inclined, but given they are simply human, and humans are simply animals, and animals, well, we can be lazy, egotistical, and what not....
What's particularly interesting is that they will ask what you are, but don't ask what you mean by that. Are you only picking one thing, because it's inconvenient to rattle everything off, or because you're ashamed of some of your ancestry, or maybe you simply don't know, and just put what you've either been told, or what people assume that you are? Maybe you just put what you feel?
So, on a very personal level, I have my doubts, but seriously, it's not a good source for this. Not at all, because it's just not very robust.
Again, they are using the very same strategy to collect economic data that they use to obtain whether you think you're related to Geronimo, or George Washington.
I think that I should have to say it's crap. I just don't see that you'd loan me money based on my personal account of my finances, and certainly my own bank will not.
It might be used for other things, less important things, but the significance of what's at sake is going to create, in me, a demand for more/better evidence than what the census can offer.
T to the E to the O As stated before, would you like to suggest a different data source? Failing that, would you like to provide evidence that the US Census data is crap?
ReplyDeleteYour anecdote is, of course, just that. Can we get data, then?
So, wait. You're not going to take my word that what I said happened?
ReplyDeleteOn the contrary; I know sufficient about statistics to know that a single error tells me nothing. So, yeah, that's not a real good rhetorical device here. Try again.
ReplyDelete(and, for that matter, that bias can be corrected for.)
ReplyDeleteAlso, this guy made an assertion, like,
ReplyDelete"The sky is actually green."
Doesn't it fall upon him to prove that, not me to disprove it?
He sees a pattern, but he isn't asking what made the pattern, nor will the data provide that info to anyone who asks.
Well, it's not so much a rhetorical device as questioning why you're trust the same anecdotal evidence simply because it's plural.
ReplyDeleteSo, they ask people, over, and over, what they make, and whether they answer honestly about what they think is one issue. Whether they answer at all is another. Whether they were actually accurate, still another.
So, let me flip it, what's their process for ensuring that their economic data is accurate?
I expect they don't have one, because the census is just them asking people how many live there, what they are, what the household income is, and so on. As I said, informal to a degree that would be unbearable to anyone with money riding on it.
Again, I think it's his job to point out stronger data, not mine.
Why do I care?
ReplyDeleteBecause it's pretty clear that this is suggesting we can stop caring about the middle class, because one day, we'll all be rich!
Maybe, that's true, but he's not really put any light on the machinations of how that could be made into something reassuring for those of us who have lived experiences that run contrary to this.
When I say "those of us" I mean middle class people.
Yes, and effectively I am saying [Citation needed] to your suggestion that the census data is crap; you are saying you don't need it. So, yeah, I think we're done here until you provide citation.
ReplyDeleteHave you not taken the census?
ReplyDeleteIf you've taken the census, contrast that with applying for anything else that requires that you prove your income. It seems self evident, but I'm fine with being done.
ReplyDeleteTake care.
The census is completely self reported and requires no validation of any kind. That's just fact.
ReplyDeleteI found out, via looking into it, that the bureau applies a bit more rigor to the collection of economic data via SIPP. They probe for "more detailed" information, but I'm still not clear about whether that's just more,
ReplyDelete"Because they told us so." questions, or them asking for documentation.
They do this to 42k participants, then extrapolate that data to apply to all of us.
So, really, the process for collection of economic data is a bit more in depth, though I can not vouch for how materially significant that is, and applies to far less people than the actual census does.
It seems like citing three sources is kind of traditional, but for some reason, just the census, with it's obvious issues, is okay here. Okay, I guess.
Doesn't the fidelity of the data matter? Doesn't it matter that what was said, what the people reported is accurate? Are we just assuming that it's accurate, and why is that?
T to the E to the O At least some of the errors should even out. For example if you ballpark your gross income, some tend to say something bigger then they earn and some something smaller.
ReplyDeleteSince asking many people seems to be a valid method of research, if done correctly, I wouldn't regard the data as 'crap', just because it was generated through questionnaires. But we can still question what this data implies. I agree that it feels like the middle class is breaking up, and not by everyone becoming rich... While I cannot point to a statistic, I do believe to have read about that several times in the recent years. So, hearsay. Still.
So, I get the idea, how statistics work. What you're saying though, is that I should trust this data, because I should believe that people are going to lie, or are simply mistaken at the same rate in either direction.
ReplyDeleteI don't think that's good science, though I do understand that it's traditional to accept that as a practical limit of inquiry outside of the hard sciences.
The SIPP data sounds like it could be more reliable, but again, does "more detailed" just mean, "We asked them to answer, on the honor system, 50 more questions than other people.", or does it mean, "We proved what they said was true by comparing what was said to what is documented."?
What I may not be expressing clearly is that I don't believe there is a strong tendency, culturally, for people to under report their income, so it seems likely to me that the data would trend upwards, and the worse off people were, the more likely an upward deviation would be.
To put it a little less delicately, if this was a piece on penis size, and it said that most of the average ones grew several inches between the intervals reported, I'd hope they were using a ruler, not just asking around.
Them asking more questions about it,
"What color is it?"
"How many times do you X with it?"
"Hairy, or no?"
"Shorn, or no?",
wouldn't make me more confident that the data was accurate.
Only a ruler, and eyes on the proverbial prize, would provide that assurance.
So, did the census use a ruler?
It doesn't seem like it, and even if they did, they seem to think that if they measure just 42 people, that can be applied to a population of 300,000, and be of use. I find that hard to swallow.
Still, it's perked my interest, and I'll keep my ears open for more about this. For me, this data, raises more questions than it answers. I hope that some independent studies are performed to drill down, and find out about some of the relationships going on that produced this data.