College Pays (Some Restrictions Apply, Void Where Prohibited)

Nancy Folbre of the TimesEconomix blog leads us to the College Board’s most recent exercise in post hoc ergo propter hoc human capital reasoning, “College Pays” (PDF). Its most insightful figure, 1.6, is on page 16: “Median Earnings (in 2011 Dollars) of Full-Time Year-Round Workers Ages 25-34, by Gender and Education Level, 1971-2011.”

Figure 1.6

Using median earnings and the 25-34 bracket are good moves. Full-time, year-round workers, however is a little dubious, though that’s the only consistent source the College Board had for 1971-1993 (Condition of Education 2004a, Supplemental Table 14-1), so I don’t mind it so much.

What I do mind, though, is the median-to-median comparisons because they hide college-educated workers who earn less than the median high-school graduate. Presumably college wasn’t that great an investment for them. On the other hand, I like the percentile-to-percentile comparisons found in Figure 1.5, which shows earnings ranges for all full-time, year-round workers ages 25 and up, except it’s not the more relevant 25-34-year-old bracket that the College Board used in Figure 1.6. Our loss.

Figure 1.5

For the record, at least 20 percent of 25-34-year-old college graduates who had any earnings (not just full-time, year-round workers) earned less than the median high school graduate the same age. Also, the Census Bureau omits people who earn nothing, which these days is about 12 percent of all college graduates and a quarter of all high-school graduates.

Percent Range College Grads Earning Less Than HS (25 - 34) Percent 25-34 With Zero Earnings

(Source: PINC 03, author’s calculations and did he mention that student loan debt has gone up?)

As always, the College Board’s actual findings are college pays off, if it pays off, when it pays off.


To editorialize on Folbre’s response, she correctly points out that the median college graduate today earns less than in 2000, to say nothing of 1971, and she notes that in fact many college graduates are simply displacing high-school graduates in the job market. Her metaphor is very similar to what I had in my mind:

Many college graduates are simply displacing less-educated workers from the jobs they once held, scrambling up the attic stairs to the roof of a bungalow whose first floor, inhabited by mere high school graduates, is now largely underwater.

I’d characterize it as people clinging to a ladders that’re sinking into the ocean, and the energy spent jumping from the “high school” ladder to the “college” ladder just to remain at the same altitude equals the rents transferred to universities.

Although Folbre isn’t a cheerleader like the College Board, her explanation for poor college payoffs is a bit iffy. To her, U.S. college-educated workers were hit with a supply shock of similar college-educated workers overseas, e.g. China quintupling the number of bachelor’s degrees it was conferring between 1999 and 2005. As the charts above show, however, plenty of college-educated workers haven’t been doing well for a while, which raises questions about the human capital hypothesis.

Trade doesn’t explain why 20 percent of high-school graduates between 25 and 34 have been in zero-earnings territory since the dot-com bust, much less 25 percent now. That’s obviously a lack of demand in the economy that’s best explained by the trade deficit rather than trade itself. Our trading partners are job-destroying neo-mercantilists, not comparative-advantage Ricardians.

Folbre also isn’t willing (in this post at least) to acknowledge other U.S. underconsumptionist policies of rewarding unearned incomes while taxing earnings, which is another monster problem that keeps workers chronically underemployed regardless of their education.

“College Pays, Sort Of” Indeed.


    1. Mo, the article for you is here. “Peak associates” largely coincides with the stagnation in the legal services sector. For that reason, I’m dubious that fewer graduates means remarkably better employment outcomes. Sure, it’s tautological that the people who don’t go to law school aren’t going to be unemployed afterwards, but (a) there are still a lot of unsustainable outcomes for law grads out there these days, and (b) I really don’t think that employers are going to enthusiastically start hiring from low-ranked schools. It’ll be a real test of the system if they don’t.

  1. Awesome quote alert: “Our trading partners are job-destroying neo-mercantilists, not comparative-advantage Ricardians.” Although I think you are aiming much more at China than say a Germany with this.

    Also this little gem rears it’s head everywhere as a cost of unintended consequences: “…rewarding unearned incomes while taxing earnings, which is another monster problem that keeps workers chronically underemployed regardless of their education.”

    Basically, it’s better for our bottom rung of society to live off a mixture of public support benefits than to strive towards anything greater, because it’s not going to get any greater with any labor they add. They have determined “The Juice is Not Worth the Squeeze” since even if they do put work into something, they will continue to be marginalized into the future (Why be a long term McDonald’s Burger Flipper?). We are taking the “Heart” out of our own poor people, because they have nothing else to believe in, and now fully recognize their own worthlessness. It’s sad. What would be considered a “Bad” decision by some is a “Great Idea” for the poor, this recent Atlantic article explains:

    1. KO, I am aiming at China, though looking at the most recent data we have a trade deficit with Germany (#4), but it’s not a developing country. China is #1, Japan #2, Mexico #3, Saudi Arabia #5 (probably all oil), Canada #6, Ireland #7, Italy #9, and Russia #10.

      Re: poor people and incentives, that Atlantic article is interesting, but there was a report by the CBO—I don’t know if I buy its thinking—that found that at low incomes (~$20,000) because they’re losing their various government benefits, poor people pay higher effective marginal tax rates than higher-income earners (PDF). How’s that for distorting incentives?

  2. The author of the Georgetown study she links to, Anthony Carnevale, is the biggest college is king cheerleader out there. The projections he and his cohorts put together on labor demands are astounding–astounding in that their methodologies are highly dubious because they basically assume market supply of college degrees is commensurate with market demand for those degrees (I mean, why would someone enter college without knowing what the demand for their degree will be in four years).

    1. Funny story: Last weekend, some newspaper ran the Obama administration’s response to the “forgive student loan debt to stimulate the economy” petition, even though that response was more than two years old. In it, the White House said that more than 60 percent of the jobs created over the next decade would require a college degree. It didn’t give a source, but the only study I could find that gave such an astonishing finding? Written by Anthony Carnevale’s Georgetown outfit based on his analysis of CPS data.

      For the record BLS’s most recent employment projections estimate about 30 percent of the jobs created by 2020 will require college degrees, and there will be more than enough graduates to fit those positions.

      1. This is the bit from Carnevale’s 2010 paper, Help Wanted: Projections of Jobs and Education Requirements Through 2018:

        Click to access FullReport.pdf

        The Georgetown University Center on Education and the Workforce has developed an approach to projecting educational demand that differs fundamentally from the BLS approach. A detailed methodology paper can be found online at In short, our method:

        “Builds on the occupation projections data provided by BLS;

        Uses existing labor market data for each year between 1983 and 2008 to analyze increases and decreases in each occupational group;

        Uses occupationally specific data on increases in education and training required by an occupation;
        Assumes that the present distribution of education among the employed prim-age population is the best single indicator of present demand for education. This model does not start with a deficit of some 20 million post-secondary degrees, as the BLS model does;

        Has proven itself to be robust across a number of statistical tests. The most intuitive of these shows that by using data between 1983 and 2002 the projections method forecasts educational demand for 2008 reasonably well.”

        For whatever reason, Carnevale’s work has clout among policymakers here in Maine.

      2. Yeah, that was the paper. The insanity I’m referring to begins on page 13. In figure 2.1 it says that between 2008 and 2018, 17 million jobs will require a high school education or less. That’s HALF of what the BLS says for 2010-2020. About 16.8 million jobs will require something more than high school, Carnevale, et. al. find 30 million.

        I’m not saying the BLS’s methodology is highly accurate and precise, but Carnevale has four years and three weeks to find jobs for a lot of college grads.

  3. ML,


    This is a rocking article…

    “Also, the Census Bureau omits people who earn nothing, which these days is about 12 percent of all college graduates and a quarter of all high-school graduates.”

    Excellent stuff – would love to know how *including* the zeroed affects the median.

    Btw, which institutions’ shady reporting practices does “zero exclusion” medians sound like – L*w Sch**ls.

    And, in today’s “Sign of the Pending Apocalypse” the ABA gets around to proposing Law School placement stat audits (only a couple of decades late).

    Personally, I think the next step is Truth, Reconciliation, and Decapitation Committees auditing *historic* placement reports…

    “Our trading partners are job-destroying neo-mercantilists, not comparative-advantage Ricardians.”

    Another gem…a *lot* (not all) of the US’ competitive problems stem from China’s long-standing policy of more or less sterilizing their huge trade deficit by means of capital controls –

    Step One – Export a ton of stuff to US.

    Step Two – Receive dollars

    Step Three – Limit ability of Chinese savers to use said dollars to purchase foreign goods.

    Step Four – Maintain artificially low Yuan exchange rate to repeat steps One, Two, and Three.


    Step Five – Employ Jedi Mind Trick to convince US government (year after year) that China doesn’t manipulate its currency.


    This is how China has amassed $1.5+ trillion in US Treasuries, 3.5 trillion in total foreign reserves, and grown the fastest economy in human history.

    While there is little question broad swathes of US industry are less competitive than Chinese industry, Chinese currency controls (and resultant FX peg) have played a big part in China’s ability to continually achieve ever higher trade deficits with the US.

    Had the Yuan been allowed to rise to market rates (by letting Chinese savers spend wherever and however they pleased) more US goods would have purchased with the more valuable Yuan – helping to offset the US trade deficit with China.

    But apparently the US government got it into its collective “head” that it was being “really clever” by accepting actual, physical Chinese goods in exchange for mere pieces of paper (whose value the US government could control at will by means of interest rate manipulation)..

    On these terms, the US government was apparently happy to let the Chinese government manipulate in *its* own way

    We get goods for paper! Tee Hee Hee Hee!

    I imagine the thought might have been that the US would continue its historic happy march to a 100% service based economy (hairdressers and public phone sterilisers of the nation, rejoice!).

    Let China do the dirty work – we’ll give them paper promises and they’ll send us sh*t.

    Meanwhile, the “freed up” US labor will all become biochemists, aeronautical engineers, and environmental sports lawyers (relying upon the scrupulously‎ accurate, market clearing employment placement reports issuing forth from America’s heavily administrator and rock-wall endowed universities).

    But the Powers that Be apparently forgot to follow through/monitor the back half of the Millennium (literally…)

    1. cas127,

      Excellent stuff – would love to know how *including* the zeroed affects the median.

      It doesn’t have that much of an effect, but the earnings are lower. If I had my druthers, I’d look at men’s earnings since they’re “not supposed to be unemployed.”

      Step Five – Employ Jedi Mind Trick to convince US government (year after year) that China doesn’t manipulate its currency.

      Naw, mostly it’s Wal-Mart and Wall Street that does the convincing. Kudos for pointing out that Chinese workers don’t benefit from Chinese monetary policy. A lot of people miss that.

      We get goods for paper! Tee Hee Hee Hee!

      Pretty much. Dean Baker likes to joke that the PBOC should’ve known that its dollar hoard would lose value at some point. See, the thing about neo-mercantilism is that you can’t control the value of the dollar unless you’re the Fed. The PBOC is not the Fed.

  4. ML,

    1) Impact of “Zero Incomes” on reported median Income – I do wonder if the excluded “zero incomes” might have a more-than-expected impact (not *huge* but perhaps significant) on the median, given how medians are constructed.

    If you have 100 rank ordered incomes and you go from having 5 to 10 being “uncounted” because they are zero (inventing numbers here for illustrative purposes) the move in the median might be marked.

    Granted, the Law of Large Numbers *might* keep things fairly stable in a universe of 115 *million* households…but the “zero income” households scale up as well (from 5 to 10 million, to use similar invented numbers).

    To the extent that the rank ordering in the middle range is a bit discontinuous (a big guess, I’ll admit), excluding “zeroes” might hike the reported median from something like 47k (“rank” speculation on my part, as it were) to the more commonly reported 50k.

    Not a *huge* difference but it would push the “2012 incomes are as low as 19xx incomes” comparisons back a number of years.

    The whole zero/median debate is very interesting because it played a very central role in how law schools manipulated the sh*t out of their reported placement salaries.

    Huge chunk of your grads not reporting salary data?

    Then they don’t exist for purposes of the median that will be published to the world (with very fine print or no disclosure of said fact).

    Adverse selection?…hmm, we seem to be running late for the Environmental Sports Law class we teach on the gold-plated rock wall…

  5. 2) Jedi Chinese/Walmart Mind Trick – “These are not the trillions you are looking for”.


    I’ll go for the fact that there are a *lot* of implicated parties profiting from the immensely unstable (and decade+ long) status quo when it comes to Chinese currency manipulation/US interest rate suppression (really two sides of the same competitively debauched coin).

    But you would think that *somewhere* along the line, *somebody* in a reasonable position of authority (Fed, Fed, Bueller, Bueller?) would take a look at the huge macroeconomic dislocations going on and at least try to step out of the toxic economic pas de deux that the Chinese/US have been engaged in for over a decade.

    But doing nothing/the same thing has apparently been too easy/lucrative and the assembling ruin too asymptomatic (Bubble? What Bubble? Inflation? What inflation?) for any new action to really be taken.

    On the US side, it has just been too damn easy to stay hooked on the “import goodies for free” train (since we have an “eternal” reserve currency whose interest rate we can zero out at will and who the hell else are the Chinese going to sell to, Bwaa Ha ha HAAA!).

    Except I think, you know, an *adult*, would realize that the Chinese are almost certainly getting the better long-term end of the bargain.

    Their physical productive capacity has soared – ours has shriveled (but our hairdresser/public telephone steriliser ranks have swelled!).

    Yep, hundreds of millions of very hard working Chinese are getting rather massively screwed on their “savings” (with the Chinese leadership operating under the “principle” – “All your savings are belong to us”) but tens of millions of 25 to 54 year old Americans are getting screwed out of a career.

    (Btw, using St. Louis Fed FRED numbers, I’m showing the US *20 million* jobs below 1970 to 2000 trendline, as of today…)

    And what happens when 200 million enraged Chinese savers realize that the US has been de facto (or, eventually, *in fact*) defaulting on the backing to their very hard earned savings?

    “Tee hee hee – free goodies” is going to turn into “Holy sh*t!” very fast if the Chinese leadership is able to convince their expropriated, pissed off, ravening masses that it is really the f***ing Americans fault (conveniently omitting their own instigation/perpetuation of the currency peg catastrophe).

    And US ruin doesn’t even have to come in the form of 200 million justifiably enraged Chinese factory workers pouring out of 2 million cargo containers in the Port of LA.

    They can just stop “selling” us sh*t on (bad, very bad) credit.

    Physical goods supply in the US will fall by 40% to 50% in short order (my wild a** guess) in an environment where Quantitative Easing Googleplex will have increased the US “money” supply to Weimar levels?

    And who can the Chinese sell to if not the US (bwa ha ha HAAA!) – how about *themselves*?

    Until very, very recently the US remained pretty damn rich doing so (intl trade being a very small % of US GDP).

    I’m a free trader (and very aware of the equally luxuriant dangers of protectionism), but in a universe where the two most powerful economic governments in the word are locked in mutual manipulation, the outcomes of pseudo-free trade are *not* going to be like in the textbooks.

    (Bwa ha ha HAAA! turns into Boo Hoo Hoo Hoo).

    1. (Btw, using St. Louis Fed FRED numbers, I’m showing the US *20 million* jobs below 1970 to 2000 trendline, as of today…)

      I thought that too a while back, but it’s just women entering the workforce. We’re short I think 10 million jobs, not 20 million.

      1. I’ll buy 10 million jobs “short” (after backing out post-1970 women’s entry into the workforce in greater percentages).

        Still a *lot* of jobs (and very roughly consistent with the 6% fall from 82% to about 76% in terms of the 25-to-54 employment-to-population ratio).

        And human capital mis-allocation due to cooked placement reports certainly isn’t helping.

  6. “and at least try to step out of the toxic economic pas de deux that the Chinese/US have been engaged in for over a decade.”

    Or at least be open and honest about what has really been going on.

    The US leadership can’t even muster mere honesty (let alone policy success) anymore.

    Which brings us nicely back to the ethical (and placement) failings of law schools…

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s