Lecture 2 - Stanford University

Report
LABOR TOPICS
Nick Bloom
Polarization and inequality
Nick Bloom, Stanford University, Labor Topics, 2015
1
Polarization
Top 1%
Nick Bloom, Stanford University, Labor Topics, 2015
2
ALM (2003) predicts labor market polarization
Goos and Manning (2008, RESTAT) point out in the paper “Lousy and
lovely jobs” that routine tasks are most substitutable by computers
Hence, occupations with a high input of routine tasks should expect to
see the largest falls in employment and income from computerization
They test this using both US and UK data and find supportive evidence
Nick Bloom, Stanford University, Labor Topics, 2015
3
First they establish that routine tasks are
most concentrated in median paid jobs
Source: Goos and Manning (2008, RESTAT)
Nick Bloom, Stanford University, Labor Topics, 2015
4
They then show median jobs have shrunk,
with a J-curve across all (UK) jobs
Source: Goos and Manning (2008, RESTAT)
Nick Bloom, Stanford University, Labor Topics, 2015
5
The jobs that have expanded are either nonroutine manual or non-routine cognitive
Source: Goos and Manning (2008, RESTAT)
Nick Bloom, Stanford University, Labor Topics, 2015
6
And in fact many low skilled jobs have grown
And that’s despite relabelling (i.e. “cleaners” now being called
“sanitation consultants” or some other ridiculous name)
Source: Goos and Manning (2008, RESTAT)
Nick Bloom, Stanford University, Labor Topics, 2015
7
These changes are within and between
industries
Source: Goos and Manning (2008, RESTAT)
Nick Bloom, Stanford University, Labor Topics, 2015
8
The Acemoglu and Autor paper shows similar
polarization by occupation (US data)
Source: Acemoglu and Autor (2010, HLE)
Nick Bloom, Stanford University, Labor Topics, 2015
9
In fact this polarization is global in scope
Source: Acemoglu and Autor (2010, HLE)
Nick Bloom, Stanford University, Labor Topics, 2015
10
Recent Autor and Dorn paper argues clear
service/non-service difference in low skill
Source: Autor and Dorn (2013, AER)
Nick Bloom, Stanford University, Labor Topics, 2015
11
Polarization of US incomes too (1/2)
Source: Guvenen, Ozkan and Song (2013)
Nick Bloom, Stanford University, Labor Topics, 2015
12
Polarization of US
incomes too (2/2)
Source: Autor and Dorn,
(2013, AER)
Nick Bloom, Stanford University, Labor Topics, 2015
13
Polarization
Top 1%
Nick Bloom, Stanford University, Labor Topics, 2015
14
Related topic is the change in inequality
over time
• Well known paper by Piketty and Saez (2003, QJE) looks at
income inequality over time, focusing on top 1%
• They do this by assembling a variety of IRS related sources
on income and estate taxes paid, plus some other
supportive data on CEO pay etc. (administrative data)
• While this is not closely related to SBTC, there are some
parallels (and I think it’s another important topic)
Nick Bloom, Stanford University, Labor Topics, 2015
15
First result is incomes are very skewed!
Source: Sept. 2013 Update to Piketty and Saez (2003)
Nick Bloom, Stanford University, Labor Topics, 2015
16
Cross sectional income distribution
Nick Bloom, Stanford University, Labor Topics, 2015
17
Second, recent income growth is also
skewed (so skew likely to keep rising)
Nick
Bloom,
Stanford
University,
Topics, and
2015 Saez
Source:
Sept.
2013
UpdateLabor
to Piketty
(2003)
18
Third, this inequality (at least for top income
shares) has a U-shape over 20th Century
19
Nick Bloom, Stanford University, Labor Topics, 2015
Source:
http://eml.berkeley.edu/~saez/lecture_saez_chicago14.pdf
Third, this inequality (at least for top income
shares) has a U-shape over 20th Century
20
Nick Bloom, Stanford University, Labor Topics, 2015
Source:
http://eml.berkeley.edu/~saez/lecture_saez_chicago14.pdf
Reason fall in capital incomes from WWII
onwards, with rebound only via wage income
Early 20th century super rich collected dividends, probably from fortunes
accumulated in 19th Century. By end 20th century more wage incomes
Nick
Bloom,July
Stanford
University,
Topics, and
2015 Saez
Source:
2010
UpdateLabor
to Piketty
(2003)
21
Reason fall in capital incomes from WWII
onwards, with rebound only via wage income
22
Nick Bloom, Stanford University, Labor Topics, 2015
Source:
http://eml.berkeley.edu/~saez/lecture_saez_chicago14.pdf
Similar results for wealth (not incomes)
23
Nick Bloom, Stanford University, Labor Topics, 2015
Source:
http://eml.berkeley.edu/~saez/lecture_saez_chicago14.pdf
The UK pattern is similar to the US
Nick
Bloom,Atkinson
Stanford University,
Source:
(2003) Labor Topics, 2015
24
Canada is similar too
Nick Bloom, Stanford University, Labor Topics, 2015
25
Source: http://eml.berkeley.edu/~saez/lecture_saez_chicago14.pdf
As always the French have to be different
Nick Bloom,
Stanford
University, Labor Topics, 2015
Source:
Atkinson
(2003)
26
Japan and Sweden similar to France
27
Nick Bloom, Stanford University, Labor Topics, 2015
Source:
http://eml.berkeley.edu/~saez/lecture_saez_chicago14.pdf
So what explains these long-run moves in
income inequality?
• Piketty and Saez (2003) speculate one factor is much
higher levels of wage and capital income tax since WWII
– Slows down the accrual of multi-generational fortunes
• Another factor could be regulation of top pay (e.g.
particularly during WWII) combined with social-norms
(e.g. after WWI greater power of unions and left wing).
• SBTC may also be a factor, although hard to square with
different French experience (which had no wage income
recovery in late 20th Century).
Nick Bloom, Stanford University, Labor Topics, 2015
28
An alternative
and interesting
take (from Greg
Mankiw’s blog)
Nick Bloom, Stanford University, Labor Topics, 2015
29

similar documents