Everything Is Bullshit: The Greatest Scams on Earth Revealed (13 page)

But this past April, a group of researchers from UMass Amherst
revealed a flaw in the Reinhart-
Rogoff
paper. Like
many economists, the researchers had been trying unsuccessfully to replicate
Reinhart and
Rogoff’s
findings. Only when the Harvard
economists shared their original dataset and Excel spreadsheet did the UMass
team discover why no one could replicate the findings: the economists had made
an Excel error.

Reinhart and
Rogoff
forgot to include
5 cells of data. Noting this mistake, and the exclusion of a number of years of
high debt growth in several countries and a weighting system that they found
questionable, the UMass team declared that the effect disappeared. Instead of
contracting 0.1%, the average growth rate of countries with debt over 90% of
GDP was a respectable 2.2%. The mistake was caught, but for 2 years the false
finding influenced policy-makers and informed the work of other economists.

Bad
Incentives

Moving
to open access journals would expand access to scientific knowledge, but if it
preserves the idolization of the research paper, then the work of science
reformers is incomplete.

Reformers argue that the current journal system slows down the
publication of science research. Peer review rarely takes less than a month,
and journals often ask for papers to be rewritten or new analyses undertaken,
which stretches out publication for half a year or more. While quality control
is necessary, thanks to the Internet, articles don’t need to be in a final form
before they are shared. Michael
Eisen
, co-founder of
the Public Library of Science, also notes that, in his experience, “the most
important technical flaws are uncovered after papers are published.”

People celebrate the discovery of new drugs, theories, and
social phenomena. But if we conceptualize science as crossing out a list of
possible hypotheses to improve our odds of hitting on the correct one, then
experiments that fail are just as important to publish as successful ones.

Journals could not remain prestigious, however, if they
published litanies of failed experiments. As a result, the scientific community
lacks an efficient way to learn about disproven hypotheses. Worse, it
encourages researchers to cherry pick their data and express full confidence in
a conclusion that the data and their gut may not fully support. Until science
moves beyond the journal system, we may never know how many false positives
are produced by this type of fraud-light
.

A
Scientific Process for the 21st Century

Although
scientists are the cutting edge, there are many examples of missed
opportunities to make the scientific process more efficient through technology.

In our conversation with Banyan, a startup whose core mission is
open science, CEO Toni
Gemayel
revealed just how much
low hanging fruit is out there. “We want to go after peer review,” he says.
“Lots of people still print their papers and [physically] give them to
professors for review or put them in Word documents that have no software
compatibility.” Banyan recently launched a public beta version of its product
— tools that allow researchers to share, collaborate on, and publish
research. “The basis of the company,”
Gemayel
explains, “is that scientists will go open source if given simple, beneficial
tools.”

Physicist turned open science advocate Michael Nielsen is an
eloquent voice on what new tools could facilitate an open culture of sharing
and collaboration. One existing tool that he advocates is
arXiv
,
which allows physicists to share “preprints” of early drafts. This facilitates
feedback on ongoing work and disseminates findings faster. Another practice he
advocates — publishing all data and source code used in research projects
alongside papers — has long been endorsed by scientists and could be
accomplished within the journal framework.

In his essay “The Future of Science,” Nielsen also imagines new
tools that don’t yet exist. A system of wikis, for example, that allow
scientists to maintain perfectly up to date “super-textbooks” on their field
for reference by fellow researchers. Or an efficient system for scientists to
benefit from the expertise of scientists in other fields when their research
“gives rise to problems in areas” in which they are not experts. After all,
even Einstein needed help from mathematicians working on new forms of geometry
to build his General Theory of Relativity.

But none of these ideas are likely to take off on a mass scale
until scientists have clear incentives to contribute to them. Since publication
history is all too often the sole metric by which a scientist’s work is judged,
researchers who primarily assemble data sets for others to use or maintain a
public wiki of meta-knowledge will not progress in their careers.

Addressing this issue,
Gemayel
references the open spirit amongst coders working on open-source software.
“There’s no reward system right now for open science,” he says. “Scientists’
careers don’t benefit from it. But in software, everyone wants to see your
GitHub
account.” Talented coders who could make good money
freelancing often pour hours of unpaid work into open-source software, which is
free to use and adapt for any purpose.

On one hand, many people do so to work on interesting problems
and as part of an ethos of contributing to its development. Thousands of
companies and services would simply not exist without the development of
open-source software. But coders also benefit personally from open-source work
because the rest of the field recognizes its value. Employers look at
applicants’ open-source work via their
GitHub
accounts (by publicly showing their software work,
GitHub
accounts can effectively function as a resume), and people generally respect
the contributions people make via open-source projects and sharing valuable
tips in blog posts and comments. It’s the type of open pursuit that you would
expect in science. But we see it more in Silicon Valley because it is valued
and benefits people’s careers.

Disrupting
Science

"The
process of scientific discovery – how we do science – will change
more over the next 20 years than in the past 300 years."

(Michael
Nielsen)

The
current model of publicly funding research and publishing it in academic
journals was developed during the days of Isaac Newton in response to 17th
century problems.

Beginning in the 1960s, private companies began to buy up and
profit off the copyrights they enjoyed as the publishers of new scientific
knowledge. This has caused a panic among cash-strapped university libraries.
But the bigger problem may be that scientists have not fully utilized the
Internet to share, collaborate, and invent new ways of doing science. The
impact of this failure is “impossible to measure or put an upper bound on,”
Gemayel
tells us. “We don’t know what could have been
created or solved if knowledge wasn’t
paywalled
. What
if Tim Berners-Lee had put the
world wide web
behind a
paywall
. Or patented it?”

Advocates of open science present a strong case that the
idolization of publishing articles in journals has resulted in too much
secrecy, too many false positives, and a reduction in the rate at which
scientific discoveries are made. Only by changing the culture and incentives
among scientists can a system of openness and collaboration be fostered. The
Internet was created to help scientists share their research. It seems overdue
that scientists take full advantage of its original purpose.

11.

IS COLLEGE WORTH IT?

A
merican
colleges and universities are enduring a crisis of faith among the public.
Although a college degree has long been heralded as a ticket to the middle or
upper class, the cost of college has increased faster than the price of health
care, housing, or just about anything else over the past 30 years. Tuitions,
student debt, and student loan default rates have all skyrocketed, leading
indebted graduates to malign their degrees and pundits to argue that college is
not worth the price tag.

Despite the pessimism, research on the financial value of a
college degree all concurs: A bachelor’s degree is a sound and increasingly
valuable investment. The extra income graduates earn more than compensates for
the high cost.

College should not necessarily be justified strictly by its
financial payoff. But appeals to the value of a liberal arts education or
cultivating civic virtue and lifelong friendships are moot for all but a
wealthy minority if a college education is prohibitively expensive.

How do we resolve this paradox that college is a sound financial
investment, yet an increasing number of students find themselves unable to pay
back their loans? When it comes to educating students and preparing them for
careers without indebting them, how do we grade American higher education?

College may still, on average, be a worthwhile investment. But
for American colleges and universities, a ‘D’ is still a passing grade.

The
Supersizing of American Colleges

Over
the past three decades, American colleges and universities have been
supersized. More students are paying higher tuitions — financed by more
financial aid and more debt — to attend colleges with expanding budgets
that fund more programs and more facilities.

Historic tuition prices offer a stark reminder of escalating
costs. If tuitions and fees had only increased at the rate of inflation since
1987, today’s smartphone toting
Millennials
would pay
$25,268 to attend Yale instead of $42,300. Amherst College students would pay
$24,741 instead of $44,610. Michigan residents would pay $5,815 at the
University of Michigan instead of $13,819.

Despite skyrocketing tuitions, many college administrators can
still look parents in the eye and say that college is not crazy expensive.
Adjusted for inflation, the amount of financial aid provided by the federal
government in the form of grants and loans increased from $12.5 billion in 1982
to nearly $170 billion today, and selective colleges offer enough grant money
that only a minority of wealthy families pay full sticker price. Between 1995
and 2007, net prices, or the price students paid after accounting for financial
aid, increased 20% at four year public schools (compared to a 56% increase in
full tuition) and 27% at four year private schools (compared to a 39% increase
in full tuition). That’s a more modest increase, but still substantial for an
already expensive service.

One of America’s greatest concerns is that college now pushes
too many students to take on an unsustainable amount of debt. Outstanding
student debt stands at $1 trillion, the result of an explosion of debt over the
past 10 years. Surveys of former students by the Federal Reserve Bank of New
York show that average debt per student increased from under $15,000 in 2004 to
$25,000 today, while the percentage of students defaulting on their loans
increased from 9.5% to over 17%. While it could be comforting to simply blame
the explosion of debt and delinquency on the recession, the upward trends began
before the economy imploded.

So why has college become so expensive?

Experts debate a number of potential reasons. Cutting edge
research centers have become increasingly expensive. (Think linear accelerators
instead of microscopes.) Colleges also compete for professors and
administrators in a labor market of well-educated professionals that is
increasingly productive and
well-compensated
, which
drives up salary costs.

But a primary reason — and in our opinion, a very
convincing one — is that colleges have no reason not to spend more and
raise tuition prices. After all, few people shop for "bargains" in
higher education; an expensive price tag typically indicates prestige. That
leaves one clear strategy for a school seeking higher standing: fundraise,
increase tuition, and spend, spend, spend.

Each year, schools compete for a limited pool of talented
students who — especially at the most selective schools — are
fairly insensitive to price. What’s another $2,000 a year for your daughter’s
future and the best four years of her life? This leads to a spending race to
woo wealthy, accomplished students with big name faculty and fancy buildings.
Since all that matters is a university’s prestige relative to its peers,
everyone needs to constantly raise funds and spend to outpace other schools.

At some point, colleges should hit a wall when not enough
students can afford the high price tag — especially at non-Ivy schools.
But since the government wants education to be affordable, the amount of aid
available has increased in step with tuitions. As subsidies increase, colleges
can afford to raise tuition and spend money on more programs, buildings, and
administrator salaries, effectively capturing the subsidy.

As a result, the default over the past decades has been spending
more to get more, not doing more with less. Last year, the president of Ohio
State University told the New York Times, "I didn’t think a lot about
costs. I do not think we have given significant thought to the impact of
college costs on families."

Worse, students may be paying more for an education that
delivers less and less core value. In their book Academically Adrift, two
sociologists find that class and study time at colleges has dwindled well below
40 hours a week and that 36% of students gain no critical thinking skills
during college.

All these fact suggest a simple narrative: Colleges have used
more and more easy money to build nicer buildings, launch new programs, and pay
administrators higher salaries, to the point that the cost of college has
outweighed the returns. The perception of college as a golden ticket is a
mirage; the reality is underemployed graduates mired in debt. But is this
narrative true?

The Case
for College

With
so many graduates struggling with debt, intuitively it seems that a number of
schools must churn out graduates who don’t earn enough to cover the cost of
their education. Yet looking at data and research on the prospects of college
graduates is unambiguous: College is a sound financial investment that offers excellent
returns.

Every year, a company called
Payscale
draws on its national data set of employee salaries and educational backgrounds
to rank colleges by their “return on investment.” Its analysis asks one
question: given the cost of a school’s tuition, fees, and living expenses, as
well as the amount of money one could make in the workforce without attending
college, how much more money do students make by attending a particular
college?
Payscale
answers the question by using real
data on the salary earned by graduates of each school over a
thirty
year
career.

The analysis reveals some interesting patterns. Engineering
schools offer the greatest financial payoff. Ivy League schools have the
highest tuitions but also the greatest financial returns of any non-engineering
college. In general, a degree from a research university is a better investment
than one from a smaller college. And many public schools perform well in the
analysis, allowing graduates to reap the same financial rewards as graduates of
pricier, private schools.

Most importantly,
Payscale’s
analysis
finds that college is a very sound investment. Over a 30-year career, the
graduates of a majority of schools will make at least $6,700 more per year than
the average high-school graduate. That is more than enough extra income to make
college worthwhile. In the 2013 rankings, Harvey
Mudd
,
a private college with a science and engineering focus, ranks highest. On
average, its graduates earn an extra $2.1 million over the span of their
careers, which is an annual rate of return of 8.3% on the college investment.
In the middle of the rankings, Georgia State alumni earn an extra $457,000 for
a return on investment of 4.3%. Over half of the nearly 1,500 colleges have a
return on investment of over 5%, and only 28 have negative returns. When you
account for grants and financial aid, these numbers look even better.

This analysis is not the last word on the financial value of
higher education. A generous financial aid package can turn average financial
returns into spectacular results. And the choice of major shapes graduates’
earnings as much as the school they attend: On average, engineers make $1.8
million more over their careers than someone working in the health support
sector. But the basic message appears sound: The extra income college graduates
make should more than offset the cost of tuition.

The
Payscale
analysis, however, does
not account for student debt. Most students cannot pay for their education up
front. Could loan payments eat up all the returns to college?

One way to answer this question is to look at what percent of
their salary graduates spend paying back their loans.
If a 25
year-old account manager spends 50% of her salary paying off student loans,
that is completely unsustainable.
She could never pay both her loan debt
and her ordinary expenses. But if she spends 5% of her salary, that would be
manageable. The nonprofit
FinAid
considers anything
below 10% acceptable; it considers 15% of one’s salary the maximum amount a
graduate should spend paying off loans.

If we look at how much indebted graduates spend paying off
student loans, we find that, on average, their debt is sustainable. It’s rare
for graduates to spend less than 4% of their salary paying off loans. But
graduates of the vast majority of schools spend a sustainable 4%-8% of their
salary paying back their loans. And graduates of every American college and
university on average spend less than
FinAid’s
upper
limit of 15% on loan payments.

All this analysis suggests that college graduates can afford
their education. Yet they don’t show whether the financial value of college is
declining over time. Could the value of a college degree be falling rapidly?

One way to answer this question, at least on a macro level, is
to look at data from sources like the U.S. Census. The Census asks about
Americans’ financial situation as well as their education status. By comparing
the difference in salaries between high school graduates and college graduates,
researchers can see how the returns to college have risen or fallen over time.
The surprising answer is that college has never been a better investment.

Pundits often grouch that the ubiquity of college graduates has
rendered a bachelor’s degree meaningless, but researchers examining national data
in a paper for the National Bureau of Economic Research find the opposite.
Despite the increasing number of college graduates, the value of a college
degree has only increased over time. In the early 1980s, the average bachelor’s
degree holder earned 45% more per year than the average high school graduate.
Even as the number of college graduates steadily increased, that wage premium
increased to 70% in the late 1990s and to nearly 80% today. The authors even
speculate that the supply of college graduates is too low.

The story here is that of rising income inequality in the United
States. According to one common explanation, technology has increased demand
for skilled labor, rewarding college graduates and hurting those without
college experience. So while the value of a college degree is increasing,
macroeconomic forces — not improvements within higher education —
are largely responsible.

Further, the average income of bachelor degree holders grew for
decades but stagnated over the past 10 years. This means that the growing
payoff of earning a degree over the last decade is a result of the collapse of
unskilled laborers’ wages. This makes college graduates relatively better off,
but it is hardly a rousing defense of higher education.

Still, the data all points to the financial logic of attending
college, even in the midst of the current weak economy. Among today’s young
graduates, 50% are either unemployed or working jobs that don’t require a
degree. Despite the bum deal of graduating during a recession, however, young
graduates in 2010 still enjoyed an unemployment rate of 9.3% compared to the
terrible 22.5% unemployment rate faced by Americans of the same age with only a
high school degree. The Hamilton Project, a think tank, finds that the extra
income young graduates make compared to young high school graduates has grown
over time — from $4,000 a year (adjusted for inflation) in the 1980s to
$12,000 today.

Skeptics might counter that a college degree seems worth it only
because smart, high-achieving people all go to college. Or that college’s true
value is the expensive piece of paper that signals intelligence and motivation
to employers.