The EconomistNovember 16th 2019 United States 372
1from 2021, for arrests and less serious
crimes. Michigan is next on the list.
About 19m Americans have felony con-
victions. Millions more have been arrested,
charged or convicted for a misdemeanour.
Perhaps one-in-three adults, some
70m-100m people, have a criminal record
reckons the Centre for American Progress,
a think-tank. Researchers say that eight
years after someone has committed a viol-
ent offence, or four years after they have
committed a property one, they are no like-
lier than anyone else to break the law.
Old records impose a broad cost, skew-
ing labour markets by discouraging many
from job-seeking. An estimate in 2016 by
the Centre for Economic and Policy Re-
search, a left-leaning think-tank, suggest-
ed the exclusion of ex-felons—mostly
men—from job markets cost at least $78bn
yearly in missed gross domestic product.
States also miss out on tax revenues. Re-
searchers at the University of Michigan set
out the details in a paper in March that
matched criminal histories to statewide
wage- and jobs- data scraped from Michi-
gan’s unemployment insurance scheme.
They showed that sealing someone’s re-
cord coincides with a 13% better chance of
getting a job within a year. Wages rise on
average by 25% in two years and the poorest
gain most. Recidivism was low.
A puzzle was why, despite such gains, so
few petition to clear their names says Gra-
ham Filler, a Republican state representa-
tive in Michigan. Just 6.5% of those eligible
(after a spell of staying clean) expunge their
record within five years. Fewer than 3,000
Michiganders do so yearly, from an eligible
pool of at least 500,000. The answer is
clear: it is a tedious process that can take
nine months and may cost $2,000 in legal
fees. For someone who has stayed clean for
years, it also feels shameful to return to re-
submit fingerprints and paperwork. “You
don’t want to run back to the courtroom,”
says Mr Filler. Other states can be worse. In
Utah it can take two years to seal a record.
Much better, therefore, if public records
could be wiped automatically. Technically
that’s easy. Groups like Code for America
help to plug relevant software to states’ da-
tabases. Politically it is becoming possible
too. This month in Michigan several bills
sponsored by Mr Filler passed its assembly,
with broad cross-party support. They
should be law within months, making
more crimes eligible to be expunged and
implementing automation for old records
from early 2022. Others including Louisi-
ana, New York, North Carolina and Wash-
ington will probably opt to go automatic in
the coming months. Some, like Illinois,
that are legalising marijuana are at the
same time enacting automatic clean slates
for some drug convictions. Congress is also
likely soon to consider clean-slate bills for
federal records.Why the bipartisan rush for reform?
Polls suggest 70% of voters like clean-slate
efforts, and both parties want ways to
shrink prison populations. An activist who
campaigned for this for years says Republi-
cans mostly seek economic gains from a
bigger workforce, while Democrats talk of
social fairness and not criminalising pov-
erty. Happily, the same policy suits both.
More broadly, states fret about puttingup economic and other barriers for so
many Americans with records. In recent
years 35 states and over 150 cities have
passed “ban-the-box” laws that forbid
some employers (mostly in the public sec-
tor) asking job applicants about criminal
records until late in the hiring process. Will
such changes and Pennsylvania’s new law
help Keith? He believes so, vowing he will
“show everyone I can advance”. 7W
hat do youget when you subtract the
yield on short-term government
bonds from that on longer-dated ones? A
powerful economic omen, if recent history
is any indicator. Around a year before each
of the past three recessions the yield
curve—which shows the return on govern-
ment bonds from very short durations to
very long ones—inverted. In July 2000, for
instance, the yield on ten-year Treasury
bonds dropped below that on three-month
Treasury bills; by March 2001 the American
economy had sunk into recession (see
chart). When the same thing happened in
March this year, alarm bells rang across
corporate boardrooms and political cam-
paigns. When the inversion deepened over
the summer, traders and pundits began to
speak of recession as a real possibility.
Now, however, the curve has righted it-
self. From mid-October, long-term bond
yields rose back above short ones (a move
accompanied by other bullish financial-
market signs, like rising stocks). Market-
watchers are asking: was that a false alarm?
Few economists think a yield curve in-
version itself causes a slowdown. The linkbetween the two has more to do with the ef-
fect of monetary policy on both. Short-
term bond yields go up when the Federal
Reserve raises its policy rate to keep the
economy from overheating. A drop in long-
term yields often occurs when markets ex-
pect slower growth ahead: a sign that the
Fed has tightened a step or two too many,
hitting the brakes hard enough to drag the
economy into recession.
This time around, the Fed seemed to
take the omen seriously. Over the course of
2019 it has first abandoned plans to keep
raising rates (which had been going up
since 2015), then cut its policy rate three
times, reducing the effective rate from
2.4% or so to 1.55%. The yield curve was not
the only thing on the mind of its chairman,
Jerome Powell: cuts were also a response to
a deepening slowdown in manufacturing
and a plateau in the growth rates of prices
and wages. But the central bank nonethe-
less responded faster and more fiercely to
an inversion than it usually does. If rate re-
ductions have in fact spared the AmericanWASHINGTON, DC
A recession-predicting omen sounds the all-clear. Sort ofThe economyInverse psychology
The worm danceSource: Federal ReserveUnited States, yield spread of ten-year over
three-month Treasuries, percentage points1987 90 95Recession2000 05 10 15 19-101234Inverted yield curves