Political Consequences
of Increased Long
LongTerm Unemployment
p y
Kenneth Scheve
Yale University
A il 28
April
28, 2011
U
Unemployment
l
t and
dP
Public
bli O
Opinion
i i
Why might increased long-term
unemployment
p y
matter for the p
politics of
economic policymaking?
Unemployed
may have different opinions
about policy than other citizens.
Mass opinion
p
about p
policy
y may
y be different in
settings when unemployment is high.
P li O
Policy
Opinions
i i
off th
the U
Unemployed
l
d
In bad times…January 1939, U.S. Gallup Poll
“Do you think government spending should be
increased or decreased on the following:
unemployment relief (WPA-Home relief)?”
upport Increased Spen
nding
% Su
.1
.2
.3
.4
.5
Support for Unemployment Benefits
„
„
0
„
Not Unemployed
Unemployed
Unemployment
rate is 16%
Difference is 23
percentage
points.
i t
P li O
Policy
Opinions
i i
off th
the U
Unemployed
l
d
In good times…1996,
times 1996 U
U.S.
S ISSP “Listed
Listed below are
various areas of government spending. Please show
whether you would like to see more or less government
spending in each area
area. Remember that if you say “much
much
more,” it might require a tax increase to pay for it:
Unemployment benefits.”
Support Increased Spending
%S
1
2
3
4
Support for Unemployment Benefits
„
„
0
„
Not Unemployed
Unemployed
Unemployment rate
is 5.4%
Difference is
essentially between
“spend the same as
now” and “spend
more”
Policy Opinions of the Unemployed
„
„
„
Differences in policy opinion robust to controlling
for observable demographic and political
characteristics.
Differences in opinion observed in many
countries around the world—in 1996 ISSP data
the U.S. differences are approximately
pp
y the same
as average difference in twelve other high
income countries.
S b t ti l significant
Substantively
i ifi
t diff
differences iin opinion
i i
observed for other policies such as relative
importance of inflation and unemployment in
macroeconomic policymaking.
But will these differences matter?
„
Possiblyy for p
policy
y conflict but unlikely
y for p
policy
y
outcomes which are central to current debates.
In
most countries, the combined magnitude of opinion
difference
ff
and the increased number off unemployed
is not enough to influence policy decisions.
Opinion differences between unemployed and other
citizens do not extend to many of the policy areas that
are central to current debates—e.g. taxes and
pensions.
i
But will these differences matter?
.25
.2
.15
.1
.05
Differenc
ce in Policy Opinions be
etween Unemployed a
and Not Unemployed
Opinion
p
differences between unemployed
p y and
not unemployed across issues in the U.S. in
2010 (Lu and Scheve 2010).
0
„
Unemployment Spending Wealthy Pay Higher Income Share Social Security Spending
Support for Trade Protection
Does high
g unemployment
p y
influence
aggregate policy opinions?
0
2
4
6
Unem
mployment Rate
Favor G
Greater Spending
5
8
7
10
U.S.
U
S 1982
1982-2008,
2008 NES
NES. Support
S pport for more
government services and greater spending and
unemployment.
unemployment
3
„
198219841986 19881990199219941996199820002002200420062008
year
Favor Greater Spending
Unemployment Rate
Does high
g unemployment
p y
influence
aggregate policy opinions?
0
2
4
6
Unem
mployment Rate
8
Government Jobs and Incom
me
5
7
10
U.S.
U
S 1972
1972-2008,
2008 NES
NES. Government
Go ernment should
sho ld see
to it that people have jobs and a good income
and unemployment
unemployment.
3
„
1972
1976
1980
1984
1988 1992
year
Government Jobs and Income
1996
2000
2004
2008
Unemployment Rate
Whyy doesn’t high
g unemployment
p y
influence aggregate policy opinions?
„
„
Self-interest
Beliefs about the sources of economic success
and failure—the role of effort and luck (Piketty
1995).
Inequality,
q
y, Representation,
p
, and Policyy
Change
„
„
Even if increases in long-term
long term unemployment in
the United States had a more substantial impact
on the distribution of policy opinion, the
experience
i
off the
th last
l t three
th
d
decades
d with
ith rising
i i
inequality suggests that the impact on policy
change might be modest.
Why?
Bartels
(2008) and others have argued that political
representatives are more responsive to high-income
constituent opinion than low-income constituent
p
opinion.
Dominant role played by interest groups in policy
process.
Caveat: Heterogeneity
g
y in Potential
Political Consequences