Transcript Labor Market Segmentation Then and Now: Implications for

Michael Reich
University of California, Berkeley
Capital on Trial conference
September 30, 2011




TEW 1979: Combined demand-side gaps and
cost-side profit squeezes to identify causes
of crises.
Found: Profit-squeeze explained mid1970s
crisis, mainly because of RSL.
Caveat: RSL was mainly about rising labor
share, not rising labor strength.
In current period, Rising Strength of
Management (RSM), especially since 2000.
RSM since 1980s
-- decline in union strength
-- changes in employment contract- toward
more short-term and less commitment
--changes in management incentives toward
more short term share price

What are the consequences?
(Note: will ignore globalization, deregulation of
finance, etc.)

Main question: Does RSM generate higher UE rates
and jobless recoveries?

RSM and changes in incentives to lay off workers

Pieces of the puzzle– in U.S. and Europe

Cycle and trend: estimated changes in Okun’s Law


Data and findings: In Europe, large change in cyclical
effect. In U.S., small change in cyclical component,
larger change in growth trend
Conclusions

Does RSM generate higher UE rates and
jobless recoveries?
Intuition
--Unions have declined
--LTERs are less common (Farber 2008)
-- Increase in short-term, dead-end low-paid
jobs = greater labor market dualism
--Therefore: more disposable workers and
slower productivity growth


Cost-cutting business model: In 1980s and
1990s, share prices increasingly fall less, and
rise more, following layoffs (Farber & Hallock
2009)

Consistent with disposable worker thesis of
Uchitelle 2006; Gordon 2010, 2011

But in 2000-2007, share price behavior reverts
to 1970s pattern (Hallock 2010)

Pattern since 2007?



Temporary contracts = about 10-15 percent
of all workers, 35 percent in Spain. Implies
80-90 % of all new hires.
Workers on temp. contracts are paid less, not
trained, rarely move to permanent status.
(Reich 2009)
Result: more employment variability with
business cycle– change in Okun’s Law
Figure 2 Trends in temporary (fixed-term) contracts in Europe,
1983-2010
Notes: Annual observations. Source: Bertola, Giuseppe 2009. “Labor
markets on the verge of a regulation crisis.” Vox-EU.
1986-2010
5
10
1964-1985
-5
0
Change in U.E.
-10
0
10
20
-10
GDP Growth Rate
0
10
20
To obtain the Okun’s Law cycle and trend coefficients I estimate the equation
UEit  ai  b
Yit
 i
Yit
(1)
where UEit equals the change in the state unemployment rate,
Yit
equals the growth rate of state GDP. The cyclical (short-run)
Yit
component of Okun’s Law is the estimated b. To get the estimated trend
and
of GDP growth, set UEit = 0. This condition implies
a
 Y  *

 
b
 Y 
(2)
Table 1.
Okun’s Law, 50 U.S. states, 1964 to 2010
Change in UE rate
(1)
(2)
GDP percent change
-0.198***
(0.019)
-0.213***
(0.021)
Constant
0.680***
(0.070)
0.699***
(0.059)
N
R-squared
2,253
.324
2,253
.352
State fixed effects
X
Table 2
Okun’s Law, 50 U.S. states, two time periods
1964 to 1985
Change in UE rate
1986 to 2010
(1)
(2)
(3)
(4)
GDP percent change
-0.214***
(0.024)
-0.236***
(0.026)
-0.175***
(0.017)
-0.195***
(0.020)
Constant
0.798***
(0.098)
0.965***
(0.088)
0.563***
(0.062)
0.499***
(0.047)
1028
.376
1028
.413
X
1225
.256
1225
.300
X
N
R-squared
State fixed effects
Table 3 Okun’s Law, 50 U.S. states, three time periods
Change in UE rate
1964 to 1979
1980 to 1989
1990 to 2010
(1)
(2)
(3)
GDP percent change
-0.205***
(0.027)
- 0.282***
(0.030)
-0.216***
(0.020)
Constant
0.908***
-0.104
734
0.667***
-0.073
490
0.631***
-0.045
1029
0.337
X
0.516
X
0.335
X
N
R-squared
State fixed effects
Table 4 Okun’s Law, 50 U.S. states, 1964- 2010, by extent
of union decline
Change in UE rate
Union decline >
median
Union decline
< median
(1)
(2)
GDP percent change
-0.201***
(0.031)
-0.188***
(0.046)
Constant
1.188***
(0.152)
0.304***
(0.072)
N
R-squared
State fixed effects
552
.331
506
.266
X
X
Table 5 Okun’s Law, by time period and extent of union decline
1964 to 1985
Union decline Union decline
> median
< median
(1)
(2)
1986 to 2010
Union decline Union decline
> median
< median
(3)
(4)
GDP percent change
-0.245***
(0.030)
-0.208***
(0.061)
-0.172***
(0.041)
-0.180***
(0.043)
Constant
0.587***
(0.060)
0.310***
-(0.057)
0.097
(0.061)
0.160*
(0.079)
N
R-squared
252
0.452
231
0.298
300
0.254
275
0.254
X
X
X
X
State fixed effects
Conclusions I-- Europe
RSM has generated more labor market
segmentation and led to more volatility in UE
(France- Spain comparison, Bentolila et al. 2011)
Many European labor economists propose:
Eliminate the sharp distinction between
permanent and temporary contracts, i.e.
move to the U.S. model (Boeri et al. “The
Scourge of Dual Labor Markets,” 2010)
But would the U.S. model be an improvement?
Conclusions II– RSM in the U.S.



Small effects on cyclical part of Okun’s Law
Perhaps because of an anomaly in 2009,
when credit channel freeze may have
increased unemployment (Gabe ChodorowReich, forthcoming)
Larger effect on reducing trend growth rates.
--Evident in long period of employment slack
from 2000 to 2007 and in high UE in the
Great Recession and the jobless recovery