Everything I Ever Needed to Know: The Middle School Premium

I recently wrote on the Census Bureau’s awaited annual update to its “Income, Poverty and Health Insurance Coverage” data for 2014. I editorialized on the decline in median earnings for 25-34 year-olds since the turn of the century. Today, I thought that rather than emphasize the negative, as I often do, I would point to the positive: For two decades, things have been getting better for people who don’t even finish the ninth grade. Way better.

Don’t laugh. This is serious social science.

Consider: If you think education increases earnings then you’d look at aggregate earnings of all persons (for the youth bracket, in this case) by education. You’d expect to see the aggregate earnings grow with an increasing proportion of the gains going to educated people. Instead you get this:

Aggregate Personal Earnings by Education (25-34, Both Sexes)

…Increasing proportions of the total going to the educated but no aggregate growth. This means average productivity isn’t improving; maybe the unenlightened are getting dumber or maybe educated workers aren’t really benefiting from more schooling.

How do you separate these factors when productivity and population (by education) are multiplied together and then added up to illustrate the aggregate above? In an unlikely parallel, Paul Krugman leads the way when trashing Jeb! Bush’s economic record as governor of Florida: Just compare the average aggregate growth rate to the average per capita growth rate. If the average per capita growth rate is higher than the average aggregate growth rate, then you’re looking at improving productivity rather than positive population shifts, which is bad.

We can do the same thing with education: average growth in aggregate earnings and average growth in per capita earnings.

Get ready…

Earnings Growth Rates by Education for 25-34 Year Olds (1991-, 2014 $)

Boom. Since 1991, the cumulative per capita growth rate for middle schoolers’ incomes has been 28 percent but for college grads (and above!) it’s been a scant 9 percent. In 42 years, both categories’ earnings will be identical.

Presumably, in about twenty-five years, young people will begin retarding their educations to avoid wasting their time learning ancillary subjects like geometry, science, literature, history, and—ungh—drama.

And if you thought law school scamblogs were bad, wait until you see Third Tier High School. Many a commode will be flushed that day, that’s for sure!

But aside from that massive finding, I direct your attention to the blue bars in the chart. As with Krugman’s argument on Jeb!’s governorship, they show exactly what we’ve feared all along: Any growth in college graduates’ productivity is overwhelmingly swamped by population shifts. In other words, more people are getting degrees, but overall they aren’t earning much more.

I wouldn’t read too much into the fact that dropping out of college is worse than not going in the first place. The stagnation of associate’s degree candidates, by contrast, is disturbing.

In truth, though, much of the growth in per capita earnings is probably attributable to college grads, so I wouldn’t completely discount them. However, most of that occurred in the 1990s. We’re in a very different place today.

Ponder that when your kids start the square-dancing unit in P.E.


Mellow is the Bubble

It’s been a few months since I’ve done one of these, and thanks to some overtime this weekend I haven’t had time to write, so here’s Japandroids’ “Adrenaline Nightshift.”

…And since I’m seeing a performance of Big Star’s Third this week, here’s “For You.”

Finally, here’s the best photo I took of the lunar eclipse last night.

2015-09-27 Lunar Eclipse



GAO Report: RIP High-Income IBR Deadbeats

We are alerted to the U.S. Government Accountability Office’s latest report, “Education [Department] Could Do More to Help Ensure Borrowers Are Aware of Repayment and Forgiveness Options” (here). The report asks one of the questions I’ve always had of income-based repayment plans: How much are people on them earning?

The answer, as of September 2014, is squat—even less than I would’ve guessed.

(I suspect the GAO chose September because it’s the end of the fiscal year.)

GAO Report--Figure 4 (Income)

Out of 11.2 million borrowers in repayment, 13 percent were in IBR and 2 percent were in PAYE (1.46 million plus 0.22 million). If you play with the numbers right that means about 2 percent of all IBR-plus-PAYE borrowers earned more than $80,000 annually. That’s about 30,000 people. By contrast 72 percent (1.2 million) earned $20,000 or less.

Other fun facts: One, about two-thirds of all IBR/PAYE borrowers are women, so we can predict that the REPAYE plan of the future, which will essentially require debtors’ spouses to pay their debts, will be an anti-dowry. Two, within the IBR group, 13 percent were paying the equivalent of a 10-year repayment plan, and for the PAYE people, it was only 5 percent, implying that perhaps some high-income debtors are not going to require loan forgiveness anyway. Three, only one-third of IBR borrowers went to grad school; for PAYE it was only a fifth.

The low-income finding is important because there have been some articles about how IBR and the changes to it confer vastly unfair benefits to high-income deadbeats who could repay their loans if loopholes were closed. For example, earlier in September, The Wall Street Journal shrieked about studies showing how IBR and PAYE are sops to doctors and lawyers (not M.D.s and J.D.s apparently), and my personal favorite occurred last February when The Washington Post ran an op-ed by the New America Foundation’s Jason Delisle and Alexander Holt, who argued against PAYE based on a lopsided hypothetical of a law grad who made $70,000. Thanks to the GAO study, this person was not only lucky as law grads go but also totally unrepresentative of IBR/PAYE borrowers.

So going forward, I fully expect media outlets and the NAF to report on how the changes to IBR broadly favor low-income debtors, and that there aren’t so many high-income debtors taking advantage of the system.

But what did the NAF actually say about the study? It appears to be shifting its focus away from IBR deadbeats to graduate debtors on PSLF specifically. That’s not really a topic I’m interested in exploring today, but those hoping the authors would apologize for wasting so much of our public-policy mental bandwidth up until now will have to wait. The IBR deadbeat might be dead, but I’m sure they’ll resurrect it fairly soon.

In the meantime, the NAF attacks IBR by blaming students for earning too little money. I’m not kidding. Consider their closing line:

Given that borrowers in IBR and PAYE have such low incomes and high debt levels, the plans look much more like very long-term programs for borrowers, not sources of temporary relief.

What does the NAF expect? The economy is still depressed. It won’t really recover without fiscal, trade, and labor reforms. It’s not the borrowers’ faults they don’t have high-paying jobs, nor is it IBR/PAYE’s. So what’s the solution? Making them pay more? It’s unclear where the NAF will go from here, but more debt, more education, and tougher repayment plans aren’t going to work. Given that the NAF took a wide swing and missed over the IBR deadbeats, I discourage optimism.

Speaking of pessimism for college grads, the Census Bureau has updated its “Income, Poverty and Health Insurance Coverage” data for 2014. As with last year, I won’t delve too deeply into the analysis, but here are median earnings by education level for the 25-34 bracket.

Median Earnings by Education (25 - 34)

Okay, the median college grad earned $1,000 more in 2014, but it’s still way below the peak in 2000.

Meanwhile, the percent of college grads who weren’t working is still 3 points higher than in 2008, and 6 points higher than 1997. That amounts to more than half a million college grads who could be working. Moreover, it’s noisier, but there’s been an upward trend since the 1990s in professional-degree holders who don’t work.

Percent of 25-to-34-Year-Olds With Zero Earnings by Education

The best we can say is that things didn’t get worse last year, but it’s much too soon to say things are getting better.

Dean Baker Gambles on Real Estate, Claims Land Prices Aren’t a Problem

It’s a little more complex than that, but Baker is known for provocative blog post titles, so it’s fair: On Thursday he put one up titled, “The Crisis of Too Little Land,” which is a response to Noah Smith’s Bloomberg View article, “The Threat Coming by Land.” Baker is well known for predicting the real estate bubble as far back as 2002, but notably, he used his insight to his personal advantage—which is totally justified: Ignore economists who get things right at your own peril.

In his excellent free book Plunder and Blunder (pdf), Baker writes that in 1996 he and his wife paid $160,000 for a condominium in D.C. and then sold it for $445,000 in 2004, a 158 percent increase once inflation is included. They only painted a few rooms. Consequently, Baker knows full well that Smith is referring to “land” as a production factor, not a physical quantity. (In equal rebuke, Smith refers to “land” as “capital,” which makes me want to bang my head against a wall.)

However, Baker goes on to correctly characterize Smith’s argument as “too much money going to owners of land,” but he disagrees as to the causes and implications. For one, higher land prices are a response to lower interest rates, and those have been falling since the 1980s. Two, dying cities with cheap land offer alternatives to employers (and workers) hoping to set up shop. Three, he dislikes smith’s use of the term “landlord,” when most of them are really “homeowners.” Finally, with stagnating or declining populations, the industrialized world will see lower land prices.

I’m mostly fine with the interest rate point, but I question whether significant changes in fiscal policies, especially property taxes beginning in the 1970s, are also responsible. Proposition 13 in California and S7000A in New York have done enormous damage to poorer households in those states. I’m sure the Reagan tax cuts didn’t help either, so I wouldn’t lay as much blame on interest rates.

The same goes for Smith’s use of “landlords” rather than “landowners,” which is a mannerism I see among land-value taxation advocates. I prefer “landowners” as well, but Baker’s preferred “homeowners” confuses quantity of owners with quality of ownership. Large corporations (think McDonald’s) own the most valuable land in the country, so taxing land values as Smith supports would hit the wealthiest landowners first and foremost. To the extent taxes would be shifted away from incomes (especially the regressive payroll tax), the harm to suburban homeowners wouldn’t be as pronounced. (They also can’t invert themselves overseas.) There’s also the problem of the sensitivity of land consumption to income and wealth: Wealthy people use more space than what they simply own, e.g. private golf courses, yacht clubs, etc. Thus, I’m dissatisfied with Baker’s terminology too.

Baker’s second and fourth points discuss the same effect somewhat: Populations leaving cities (or dying off) reduce land prices, but no one wants to set up shop in areas that are declining. There’s a post hoc quality to Baker’s argument here: If urban sites outside of New York, San Francisco, Silicon Valley, etc., are cheaper options, why haven’t industries moved to them already? The reason is clear: There is no good alternative to a good location, so when the government subsidizes location owners and taxes producers, our economy seizes up. This is why Smith writes, “It doesn’t matter how much empty land is out there — who wants to live on the Kansas prairie? What matters for the value of modern land is the incentive to locate close to other people.” It’s also why Baker’s title appears to miss the point in contrast to his usually innocuous tongue-in-cheek blog posts: It matters quite a bit if land-market failures have cost the U.S. 13 percent of its GDP since the 1960s.

I have two points of my own to add: One, Baker is often concerned with the U.S. trade deficit, particularly with China, but a lot of it is due to oil consumption caused by urban sprawl. Our energy policy contributes to other problems, so it’s unfortunate Baker doesn’t draw the relationship.

Two, Baker regularly touts financial transaction taxes and even taxes on vacant homes. It’s bizarre that he’s unwilling to endorse land value taxes notwithstanding his arguments against Smith. LVT is better than both.

Ultimately, Baker understands land prices in a bubble context, but he doesn’t understand land as a production factor, even though he an recite “land, labor, and capital” as a mantra (see the comments). Frankly, I’m not sure if this is worse than Smith’s land-as-capital model, but it’s highly problematic. For example, Japan has a declining population, and with it low expectations of future land rents, which in turn leads to lower inflation, low interest rates, and low growth. By Baker’s reasoning these forces should also lead to higher land prices. The virtue of LVT is taking the speculative element out of investment. It’s disappointing that Baker sees land pricing only in its excesses but not in its general effects on factor distribution and growth.

Class of 2014 NALP Data: Unemployment, Small Firm Jobs Down

A few weeks back, the National Association for Law Placement (NALP) uploaded its national summary chart that’s the basis for its Employment Report and Salary Survey (ERSS). It’s here (pdf). Two things worth noting: One, this year’s version doesn’t include the total number of graduates or the number who responded to the survey, making it impossible to determine non-responses. I have no idea why the NALP did this.

Since the class of 2013 ERSS appeared to use the same total number of grads as the ABA did that year, I’ll assume it’s 43,832 this year, which includes graduates from the three Puerto Rico law schools. Also, this year the ERSS changed 2-10-lawyer firms to 1-10-lawyer firms. I’m not sure if this is a typo or if it’s meant to separate graduates who work under solo practitioners as non-lawyers from graduates who start their own solo practices.

Okay, some analysis. Obviously this year’s ERSS confirms what’s been widely reported since the ABA’s version of the same data came out several months ago: Unemployment is down, as is the number of grads in bar-passage-required jobs, and with fewer graduates, the percentage of employed grads rose.

Here are charts of the number of graduates by employment status and the percent employed. (These exclude non-responses.)

No. Grads Employed by Status (NALP)

Percent Employed by Status (NALP)

These charts also illustrate the remarkable growth in J.D. advantage jobs over the years.

Here’s a detailed version graduate employment but with full-time and part-time status and only going back to 2007.

No. Grads Employed by Status (Incl. FT-PT) (NALP)

The question that I don’t think has been addressed is what kind of bar-passage-required jobs are responsible for the drop in that category. I won’t show all the math, but the answer is overwhelmingly private-practice, small-law firm jobs: 1/2-10-lawyer practices and solos.

No. Graduates Employed by Size of Firm (NALP)

Interestingly, the number and proportion of grads reported as starting their own practices did not change much since 2013 (-175 from 1,378). I draw two conclusions from this: One, small firms looking for new lawyers will need to look harder. I have no idea if that will push up wages in the future (there’s trivial evidence it has this year for full-time, wage-and-salary jobs). On the other hand, these could be eat-what-you-kill arrangements, which wouldn’t cost these firms much. Nominal wages for 1/2-10-lawyer practices are still way down from 2007, but the proportion of grads reporting wages is up, so this is a phenomenon to look for. The better these jobs pay, the better grads do overall: It’s the marginal graduate who matters most.

Speaking of whom, and this is my second thought, grads appear to prefer J.D.-advantage jobs and unemployment to small firm work. Given the definition of J.D. advantage, which is so broad that it likely includes graduates returning to their prior jobs, graduates’ employment “choices” don’t speak highly of small-firm work.

Thus, there is still much slack in the legal labor market, but it is improving. Big law isn’t hiring the way it used to, but fewer grads are working in smaller practices:


Cumulative Percent Change in Grads Employed in Law Firm Jobs by Firm Size (Index 2007=100) (NALP)

(Sorry this one is a little unclear.)

To conclude, this ERSS verifies the odd accounting identity explaining law graduate employment: The first people who don’t go to law school are the first ones to not be underemployed after graduating. Small beans for the thousands of unemployed grads though.

Why Debtors With the Smallest Incomes Have the Larger Problem

…Is how The New York Times meant to title Susan Dynarski’s Upshot piece, which was instead titled, “Why Students With Smallest Debts Have the Larger Problem.”

The article makes other odd or incorrect statements. For example, it says that 7 million borrowers are in default, but looking at its sources that probably includes some debtors who have both direct student loans and guaranteed student loans. (The link in the article doesn’t help.)

It then claims that debtors with high debts are less likely to default because they tend to be high-income professionals. In fact, those debtors are more likely to be sophisticated enough to sign on to a hardship deferment when trouble arises, or more commonly use IBR and its friends.

The statement also commits what’s rapidly becoming the cardinal sin of student debt reporting: using debt as the independent variable and not income. People can’t decide how much money they make, so if they have high debts, then they’re hosed. It’s not all sunshine and roses for high-income, high-debt workers either, as the interviewees from The Wall Street Journal illustrated a couple weeks ago. debtors with small balances and high incomes don’t default on their loans.

I won’t beat up on the piece too much. Its main point is that the average defaulted balance is fairly low, and given that I regularly report on that as the student debt crisis, I acknowledge that the NYT has gotten to the right place despite some staggering. It’s also correct to say that reducing balances won’t reduce defaults, as the author states.

One thing that will reduce defaults is increasing borrowers’ incomes and discouraging people who have little chance of completing college from attending. The latter policy, of course, contravenes the well-established view that everyone can get ahead of everyone else simultaneously thanks to credentials. Instead, the author argues for extended, income-based repayment, which essentially normalizes default—but we won’t call it that.

Someday the media will discover that job creation and higher pay will reduce defaults, but that day is still a long way off.