U.S Unemployment By Education Level | Hacker News Reader

U.S Unemployment By Education Level

bls.gov

91 pointsshrikant63 comments

Here's the unemployment rate data by education level in chart form:

http://www.deptofnumbers.com/unemployment/demographics/

It's interesting to observe the structural nature of these series. At no time (in the visible history of the chart) does the unemployment rate of a higher education level exceed the unemployment rate of a lower education level. Through recession, recovery and expansion, education looks to be a verifiably sound strategy for increasing your employment prospects.

Also worth noting (from the original link) is that the labor participation rate (employed + active job seekers as a fraction of the population) is higher as education levels increase. You're more likely to be working or looking for work the more education you have. Lower education levels have greater rates of people not working and not looking for work.

"Through recession, recovery and expansion, education looks to be a verifiably sound strategy for increasing your employment prospects."

That assumes a causal relationship. The chart only shows correlation.

This is the oldest, yet often useless, criticism in the world. Statisticians know that when they are observing correlations they don't have statistical evidence for causation. That requires you to take a step back, decide what other variables could cause something, and test for that.

But here it's more than some vacuous claim to correlation.

Do you get unemployed as a BA first, or do you get your BA?

Econometrics is all about identifying causation in correlations in natural data (i.e. you can't run scientific experiments on people's lives, like telling random sample of people to go to college and others to not). Here, you just need intuition to realize that one thing consistently is a prerequisite to another, to realize that it's not just some statistical garbage.

There are three possible models to explain the correlation:

1. Something influences both one's propensity to get more education (E) and not get laid off (L).

2. One's education determines one's propensity to not get laid off

3. One's propensity to not get laid off influences one's education. This is nonsense.

In a Bayesian network, you can visualize 1. as features f0,...fi (say race, parents' income, etc) influencing both (E) and (L), with (E) also influencing (L) directly. 2. would be a set of features f0,...,fi influencing (E) but not (L), with (E) directly influencing (L).

So to conclude, "This chart only shows correlation" is a useless comment to make unless accompanied by an analysis of the possible probabilistic models that the evidence supports.

> "This chart only shows correlation" is a useless comment to make unless accompanied by an analysis of the possible probabilistic models

Correlation implies causality until proven otherwise? Do you work for the government?

> Statisticians know that when they are observing correlations they don't have statistical evidence for causation.

Scientifically inspired voodoo is still voodoo. Say those with higher education have lower employment, you still don't know why. These people could be wealthier than the rest and have better opportunities in life REGARDLESS of their "education" (as one example). The macro-economist would then cite this data point is proof for needing more spending in higher-education, when in reality, this may not be the case.

Correlation implies causality until proven otherwise?

Where did I say that? I said that a comment like "this chart only shows correlation" is as useful as "The sky is blue." The sky is probably blue.

Please refer to the models that such data could support (my previous post). The joint influence of other variables like f0,...,fi is certainly discussed.

The macro-economist would then cite this data point is proof for needing more spending in higher-education, when in reality, this may not be the case.

Please refrain from judging a profession unless it is one that you are familiar with. I mentioned econometrics, which is a specialized form of statistics that focuses primarily on extracting information from observational data.

You claimed OP had no right in claiming causality could not be established unless an opposing data-set was present. Who says this data has to be present? Are econometrics infallible?

You're trying to analyze the human motivation for action in hopes of altering future action, all inside a vacuum void of real-life tests. You claim this type of empirical knowledge is impossible to attain, "(i.e. you can't run scientific experiments on people's lives, like telling random sample of people to go to college and others to not).", but I would disagree.

Companies go to great length to mine data about their customers and their behaviors, with the opportunity to run a-b tests and isolate causal relationships. We should should recognize econometrics for what it is, and that is a theoretical science, and that anyone has the right to question the integrity of claimed causal relationships.

"This is the oldest, yet often useless, criticism in the world."

And criticizing people for pointing out that correlation does not imply causation is the newest, yet often useless, criticism in the world.

Read the context of my statement again. The comment to which I was replying called it "a verifiably sound strategy". Sounds a little strong when the only presented evidence does not control for factors like parents' income (or f0, ... fi as you call them).

Maybe some such evidence does control for those factors, but the evidence presented did not.

Please read my conclusion again:

So to conclude, "This chart only shows correlation" is a useless comment to make unless accompanied by an analysis of the possible probabilistic models that the evidence supports.

I offered more than a simple criticism.

I think maybe you don't understand what correlation means. That the variables of employment and education are correlated means that a change in one variable means that the other variable will change. That is, the two variable change together. It's not necessarily casual relationship but there is a relationship.

Unemployment certainly isn't the cause of education; Thus if there was a third variable causing both low unemployment and education it must be some intrinsic quality of the individual, such as "hard work" or "intelligence".

Thus it is still good advice to tell students to work hard and do well in school, rather than saying it's okay to be lazy and not pass exams because, well, "education is only correlated with employment prospects".

What if the third variable is being black?

Or, more likely, coming from privileged families.

Not necessarily. Sometimes the better advice is to tell someone to _not_ go to school. Someone who might be a very effective (and well paid) tradesperson, but has no aptitude for University, would be much better served (and probably less stressed) by entering the workforce as young as possible, becoming a journeyman, and raising through their career without a formal education. There are at least three reasons for this:

1. By entering the workforce as young as possible, total lifetimes earnings (and more importantly, savings) are maximized for a longer period of time.

2. If the individual has no aptitude (or desire) for higher education, then there is a very good chance they will neither get the credentials, nor the knowledge that comes with a high level education.

3. Even if they did manage to squeak by and, through sheer grit and determination and stubbornness, acquire the higher education - they might now slot themselves into a field of endeavor for which they are singularly not suited for - only to find 10 years after working some white collar position, that they really, really wished they had been in construction/electrician/blacksmith/what have you...

Frequently the path to maximizing total lifetime earnings, happiness, impact, and productivity is not through higher education. I think that there is a lot of observer bias on most of our part, because most of us have gone through at least some post-high school education and would to believe it generally offers value to those who do likewise.

I've got a least one friend who really should have never gotten his Computing Science degree. He is brilliant in ways that I'll never, ever, match - and should be building houses, not trying to work in IT - just not his gig.

On the flip side, i have a quite comfortable six figure income, and never completed university - I just have an aptitude for working in IT.

I really believe there is a correlation/causation problem with the "Higher educations leads to Higher Income" - the one exception being that a Yale/Stanford/Princeton/Harvard/MIT/Ivy League graduate probably raises the ceiling (and position opportunity) for candidates above what someone without those credentials would have. But now we're talking about the difference between making $150K/year as a mid level manager and $750K/year as a Sr. VP - not as relevant to the masses.

Great points, I'm very intelligent and fairly well educated (informally) but have absolutely no use for the University system.

I did exactly that, out of high school into a warehouse job, then tech support, then software engineering, when I look at the cost of student loans, lost work, lost investment opportunities, I put the cost of my BA somewhere around $400,000.

I meant, if you're invited to a high school to talk about raising your employment prospects, it is good idea to tell students to work hard and do well.

There are definitely people who would be better off dropping out as early as possible and start working but that's a minority.

Even electricians and mechanics require tertiary education. (Otherwise, what does my brother training to be a mechanic mean when he says he's going to "class"?!)

Dropping education as early as possible might be a good idea if your aim is to work in a field where knowledge can be easily self taught. (Like IT or painting). Even then, a one year crash course I think would be quite helpful. I self-taught myself programming too, but I only learned about "unit-testing" in university. I guess you can argue I could learn that in my first job, but what if I want to, say, do a startup? ;)

>> I really believe there is a correlation/causation problem with the "Higher educations leads to Higher Income"

We're talking about higher education leads to higher chance of being employed.

I would think the "looks like" covers that.

Well, what looks like? The evidence presented doesn't control for many factors which would intuitively be likely to matter, such as parents' income.

I didn't take "looks like" as a mere hedge on the claim, because the poster clearly said "verifiably sound".

The above chart series is why I consider getting the highest education possible a categorical imperative. As much as we snark about the English grad student barista - the barista is doing better than the high school dropout who hasn't worked for years because his industry is as flaccid as a used balloon, and he has no other serious skills.

Well, who has more debt? And what if we compare her to a high school graduate employed at a similarly low level rather than an unemployed dropout?

There are PLENTY of reasons why it might not be worth it to spend as many years in education as society allows.

What guarantees that the high school drop-out's industry is "as flaccid as a used balloon"?

As a high school drop-out in the computer industry, the industry does not seem the least bit "flacid" to me.

Of course, I also had comparatively wealthy parents and an otherwise privileged upbringing, very similar to that of many individuals that comprise the lowest unemployment bracket.

More precisely, 'being from a background where you can get a higher level of education and then actually getting that education' is a sound strategy for improving your employment prospects.

Seeing this chart broken out by, say, parental income and assets would be interesting.

I'm not trying to make the absurd point that education is useless/bad, just that this is heavily conflated with all sorts of age/race/class confounding factors. The factors that might lead to outcomes like "less that a high school diploma" or "_some_ college" include things that won't be great for employment either - chaotic home life, illness or disability, requirement to care for illness/disability, being a participant or a victim in crime, drugs or alcohol abuse - or having parents that are affected by these factors.

Not wanting to be too snarky here, but a lot of posts (not yours, specifically) on inequality and social problems here on HN take on a unpleasant tone of undergraduate geek self-congratulation - "how _smart_ I am to be getting a degree with teh computers, if only all those poor people were as hard-working and clever as I am".

It never ceases to amaze me how even decades after MLK and the end of apartheid in America, that people still continually try to reframe these discussions into

"because they're browner than me they deserve to be poorer than me"

Now, I now that isn't what the OP probably intended, but (like sexism) the racism bias still runs deep in that land and is subtle and insidious.

Does the 3% unemployment rate difference between a high school graduate and a college graduate makes it automatically worth it to invest 4 years, and thousands of dollars in higher education?

Interesting that in the ethnicity breakdown, Asians have a very different line to everyone else.

I find it frustrating that those with an associate degree are counted with "some college". I'd be willing to bet the numbers are much better for people who completed an associate degree than for those who failed/dropped out of college.

It's too bad they stop with Bachelor's - does the trend continue for people with graduate degrees?

Or does it get worse? Do they get written off as "over qualified"?

I.e The boss is feeling threatened excuse.

Or too much of a "flight risk" when a higher paying position elsewhere arises, wasting the employer's investment in training.

maybe "over specialized"

over qualified is something people hear, too. Like a Ph.D. working for Starbucks. Their degree warrants a wage the company [won't] pay. They're "too good" for the alloted budget of a new employee.

The BLS doesn't report the data monthly, but it does occasionally break out the unemployment rate for higher levels of education.

http://www.bls.gov/emp/ep_chart_001.htm

In short, yes, the trend continues.

A story submitted in the last day or so

http://news.ycombinator.com/item?id=2522135

seems not to have had much discussion, perhaps because the submitter retitled the story, but the submitted story has a lot of good discussion on the value (or not) of attending college, and is worth thinking about in connection with the statistics submitted here.

Not surprising. Decent jobs where a lack of high school or college education is acceptable don't really exist anymore, and we use the checkmark of high school/no high school or college/no college as a screening device.

I saw a posting for my local city's parks department for a person who basically mows grass. Education requirements? High school diploma, associates degree preferred.

The correlation/causation issue is that we assume that no jobs for folks out of high school equates to no "low skill" jobs.

It seems to me that the real reason is that the attributes one posses to drive them through college are also valuable attributes in business success.

Someone who has no interest in working hard to attain good grades in high school and therefore does not go to college is, generally speaking, someone who has no interest in working hard to get a good job. However, if you took one of those employed college grads and stripped them of their degree, I think we would find their success in business would remain unchanged. It is the person, not the degree, that leads to employment.

Ultimately, we simply do not have enough information to reach any conclusions from the data.

I really do wish we could find a way that was cheaper than college for what is, in these cases, essentially a personality/intelligence test.

Employers used to give out aptitude tests all the time, prior to the rise of disparate impact lawsuits. Who knows how much better we could have gotten at testing these sorts of things if colleges hadn't essentially been granted a monopoly on employment psychometrics.

I believe that it is actually illegal under federal law to require education/skill levels that exceed those necessary for the job. This is civil rights era legislation, because excessively high requirement were routinely used to exclude blacks from jobs.

Educational requirements are invariably soft; I don't even have a high school diploma, but that's never disqualified me from a job.

We're in discussions regarding the purchase our small company, where I serve as (a technical) CEO. I'd be joining the purchasing company as CTO; nobody yet has asked me about having a degree in anything.

For most jobs, what educational requirements really express is the minimum level of skill in written and verbal communication, analytical abilities, and a sufficient level of proficiency in the trade in question.

When everyone in you field went to college, nobody even bothers to question whether you did, too.

Well, it sure sucks to be a high-school dropout.

But the situation is much worse than the statistics show. They dont count people who are too discouraged to look for work, those who would take a job if one were realistically available. Also not counted are those who are working fewer hours than they want. Typically the unemployment statistics underestimate by a third. I surmise, but havent dug into it, that the under-counting is correlated with education: unemployment is probably double the BLS number for the dropouts. And for those under 25, it is over 50%.

Apologies for the lack ok citations.

You can see some of that in the participation rates.

Too bad the highest education level in the survey is "Bachelor or higher". I would have really liked to see the impact of having a Master or Phd. I wonder if an education level above Bachelor makes any difference in regard of unemployment or just for income. I am also curious how these numbers looked 30 or 40 years ago.

Also interesting:

NY Times: The Jobless Rate for People Like You (2009)

http://www.nytimes.com/interactive/2009/11/06/business/econo...

Plots unemployment by race, education level, age and sex.

So what would a contractor, or an actor, or someone not actor but from the entertainment industry report? Some of those folks are regularly without job for sometimes more than a month, until they are on to a new project...

They would probably report as self-employed, although unemployed is possible.

Unemployment in government statistics is an economic-political term basically defined as "people who are able to work, who are actively looking for work and have been for the last 4 weeks"

Its important that they are "actively" searching for work. This is generally (in most countries, including Canada and the United States) defined the same as it is in the relevant welfare codes: looking at job ads, submitting resumes, making physical effort to find a new job (albeit, the extent to which that effort is made is hard to measure)

Entertainment industry people often collect unemployment between jobs, so my guess is they report as unemployed.

This apparently refers to the people registered as unemployed (and who keep registering every week), versus people who do not get a regular paycheck. I am not sure if the latter statistics exists for the US.

this also is only people aged 25 and over, not showing the tough job market for new college grads as well

What would that number be useful for? Anyway, you could just take the population of the US and subtract the "employed" number.

Depends what you are looking for. Do you want the number of people without jobs (including children, retired, hospitalized, etc.) or the number of people capable of work but without?

According to

http://www.bls.gov/news.release/empsit.t01.htm

there are 239,146 people aged 16 and over in the US who are civilians and not in institutions (like prisons or hospitals), of which only 139,674 are employed.

That leaves 99472 unemployed. Which is about a 42% unemployment rate.

That's not a particularly useful statistic. How much of that 42% is kids 16-18 in high-school, people in college 18-22, people in grad school 22-25, the retired?

In 2006, ~30 million people were enrolled in the final two years of HS, college, or graduate school: http://www.census.gov/prod/2008pubs/p20-559.pdf

Adjusted for demographic trends, that's probably higher now.

In 2009, about 12.9% of the population was over 65: http://quickfacts.census.gov/qfd/states/00000.html

Together that's roughly 70 million people who are mostly being counted in your "unemployed" figure that shouldn't be included.

That's not even including women who don't work. There is still a delta of ~13% b/w male and female labor force participation rates. They're 122 million women age 16 and over in the US. That's probably another ~10m who shouldn't be included in your unemployed figure.

The above estimates double-count: those enrolled in school who work, elderly who work, etc, but is much more realistic than your figure.

You do a good job of segmenting the population of unemployed people in to various groups, but you make not even a single argument for why they shouldn't be counted as unemployed.

For instance, you say, "women who don't work ... shouldn't be included in your unemployed figure." Why not?

You have a point there, but I wouldn't be quite so sanguine. Look at the different participation rates of graduates and those without a HS diploma. The difference is huge, and I think it tells us something about the real unemployment rate.

Another issue is that those counted as employed include a large number of underemployed, i.e people forced to work part time because they can't find more work.

In the United States, these official statistics are based on sample surveys. There are distinct survey questions about labor force participation and the more narrowly defined concept of "unemployment." Description of the methodology used in the United States is found here:

http://www.bls.gov/cps/faq.htm

and, specifically,

http://www.bls.gov/cps/cps_htgm.htm

The United States federal definition of unemployment and labor force participation are

"Who is counted as unemployed?

"Persons are classified as unemployed if they do not have a job, have actively looked for work in the prior 4 weeks, and are currently available for work.

"Workers expecting to be recalled from layoff are counted as unemployed, whether or not they have engaged in a specific jobseeking activity. In all other cases, the individual must have been engaged in at least one active job search activity in the 4 weeks preceding the interview and be available for work (except for temporary illness).

"Who is not in the labor force?

"Persons not in the labor force are those who are not classified as employed or unemployed during the survey reference week.

"Labor force measures are based on the civilian noninstitutional population 16 years old and over. (Excluded are persons under 16 years of age, all persons confined to institutions such as nursing homes and prisons, and persons on active duty in the Armed Forces.) The labor force is made up of the employed and the unemployed. The remainder--those who have no job and are not looking for one--are counted as 'not in the labor force.' Many who are not in the labor force are going to school or are retired. Family responsibilities keep others out of the labor force."

right on. isnt that what the participation rate is about? those numbers tell a much more interesting story than the unemployment numbers and correlate strongly with education level

Is this normal? Or a recent development?

Correlation is not causation, etc.

No, but robust correlation implies causation somewhere nearby. It could be A->B, or B->A, or C->A and C->B, or something more complicated. This is where common sense and ability to propose mechanisms comes in and while I can easily see how less education could make someone less employable, I don't see how someone's current unemployment can change someone's past education attainment, for instance.

I was attempting to make what I thought was a rather obvious caution (which is why I kept my post so terse), but since my above comment has been downvoted to hell[1] I guess I should explain. I was not speculating that unemployment changes education background. My common sense agrees with yours in this regard. What I was trying to remind everyone is that this data doesn't distinguish between "education makes you more employable" and "bright/hardworking/advantaged/<insert hidden variable here> people do better in school and at the work."

[1] Why is the -4 display cap still in effect now that we can't see other people's scores?

Why is the -4 display cap still in effect now that we can't see other people's scores?

I think because the current experiment of not showing comment karma scores (other than your own comment karma scores) is done with minimal changes to the existing code base. See

http://news.ycombinator.com/item?id=2515825

for more background on the current experiment.

  > Correlation is not causation, etc.

This certainly is a true and tested adage, but I've got a feeling that the correlation-backlash has gone too far in the other direction. Don't just repeat it without getting it. If substantiated through proposal of a suitable mechanism, correlation is solid corroboration of a causality-claim.

More precisely, correlation most definitely never implies absence of causation.

Upvoted because that is the most succint way I've read the response to that common retort in this entire thread.

Reply on news.ycombinator.com