Conference Papers
Upjohn Research home page
2010
Use of Administrative Data to Track Employees
and Firms: The Case of Auto Workers and Their
Communities
Randall W. Eberts
W.E. Upjohn Institute, [email protected]
Citation
Eberts, Randall W. 2010. "Use of Administrative Data to Track Employees and Firms: The Case of Auto Workers and Their
Communities." Presented at the Workforce Strategies Workforce Group, April 9.
http://research.upjohn.org/confpapers/11
This title is brought to you by the Upjohn Institute. For more information, please contact [email protected].
Use of Administrative Data to Track
p y
and Firms:
Employees
The Case of Auto Workers and Their Communities
Workforce Strategies Workforce Group
University of PEI
Randall W. Eberts
W E Upjohn Institute for Employment Research
W.E.
9 April 2010
Introduction
•
•
•
•
•
The restructuring of the auto industry and the recession has caused
considerable upheaval in many parts of the country.
For many
man regions,
regions partic
particularly
larl in the Mid
Midwest,
est the lilivelihood
elihood of displaced
auto workers and the viability of their communities are in question
For other regions, which have benefited from the restructuring, filling the
demand for more auto workers is an issue
D i i makers
Decision
k
iin b
both
th regions
i
need
d accurate
t and
d currentt iinformation
f
ti tto
address the issues pertinent to their local areas
Pose a set of questions that are perhaps more germane to a place like
Michigan, which has experienced a huge structural and we hope some
cyclical change in the auto industry
• Traditionally known as the auto capital
• Epicenter of the current restructuring in the domestic auto industry and vital importance
to Michigan’s economy
• Still claims 16% of US auto jobs, down from 28% ten years ago
•
Assess what data are needed to address these questions and give a few
examples of what we know from these data about the situation in Michigan
and mention along the way what we don’t know and would like to know
2
Basic Questions
Where have all the auto jobs gone?
•
•
•
•
Assess the magnitude and timing of the job loss.
loss Is it cyclical or
structural?
Are the losses concentrated in a few places or more widely
spread?
Are these effected regions experiencing overall employment
decline or expansion?
What are the driving industries in these regions?
•
•
•
•
Identify and examine communities that have successfully transitioned
away from auto jobs during the past decade of job decline in the auto
sector
What occupations are in demand?
Examine dynamics of the auto industry—openings
industry openings, closings,
closings
expansions, and contractions across detailed auto sectors and
establishment size
Examine the dynamics of the communities affected by the auto
displacement openings closing
displacement—openings,
closing, expansions
expansions, and contractions
across industries and establishment size
3
Basic Questions
Where have all the displaced
p
auto workers gone?
g
• Track the employment outcomes of those who were
displaced by the auto industry, focusing on their …
• Return to work either back into the auto industry or into other
industries (high-growth or low-growth industries) and in what
occupations
• Relocation to other communities
• Transition into retirement
• Participation in workforce development programs
• Earnings differentials before and after displacement.
4
Basic Questions
Where can all the auto workers go?
• Assess the potential of displaced auto workers to fill jobs
in high-growth, high-wage industries
• Examine related skill sets within occupations across industries
(using O*Net)
O*Net), and then match these skill sets with those demanded
by high-growth industries, such as “green jobs,” identified in their
local and regional labor markets.
• Track the progress of workers in returning to work and communities
in recovering from the loss of jobs,
jobs by following employment
patterns and community restructuring.
• What is region’s potential to generate jobs in highgrowth, high-performance
g
g
sectors?
• Industry cluster analysis
• Asset mapping, including human capital, physical capital, and social
capital
5
Basic Questions
What is the access to workforce development, education, and
economic development services?
What is the access to technical assistance by businesses?
How effective are these services?
For example, what are the participation rates and how effective are
reemployment services and training in assisting displaced auto
workers to regain employment?
•
•
•
Documentt the
D
th use by
b auto
t workers
k
off the
th workforce
kf
development
d
l
t system,
t
specifically the WIA Dislocated Worker Program, the Wagner-Peyser
Employment Service, and Trade Adjustment Assistant Program
Evaluate, using matched comparison groups, the effectiveness of the
workforce development programs on auto workers over the past few years
in helping them return to work and regain their prior earnings levels
Use the evaluation to inform decisions as to the best use of programs and
services to meet the needs of workers and their communities.
6
What data are needed?
• Examining
g many
y of these q
questions requires
q
flows not
snapshots
• Most readily available data are snapshots, but great strides
have been made in providing flow data
• Worker transitions:
• Employment to unemployment
• New hires, separations, accession
• Return to employment: same industry, different industry;
same occupation, different occupation
• Reemployment
p y
services;; training,
g, education
• Employer transitions:
• Openings, closings, expansions, and contractions,
relocations
7
What
at Data
ata a
are
e Available
a ab e a
and
d St
Still Needed?
eeded
• Available data:
• UI wage records
• Administrative data: workforce system, education system
• Need data:
• Employer workforce skill needs
• Worker skills and credentials
• Linked data:
• UI wage records and administrative data: workforce and
education
• Data Quality Initiative
• Quarterly Workforce Indicators (LED
(LED, LEHD)
• Geo-coded data
• OnTheMap, for example
8
What Do We Know?
• Most of what we know comes from cross-sectional
(snapshot) data
• Important information
• Need to follow workers and employers through
their transitions to answer the basic questions of
where workers go after displacement and what
• Increasingly, flow data are becoming more widely
available
• For many questions, we need to have access to the
flows and not turn them into snapshots
9
Decline in auto employment (and production)
started
sta
ted at least
east a decade ago for
o bot
both tthe
eU
U.S.
S a
and
d
Michigan
Index of Auto Manufacturing
g & Production
Index 100 = June 2000
140
120
100
80
60
40
20
0
1990
1992
1994
MI
1996
1998
2000
Rest of U.S.
2002
2004
2006
2008
Domestic Production
Source: Based on BLS CES data; BEA motor vehicles data. Note: Shaded areas represent approximate duration of recessions.
Where have all the auto jobs gone?
10
Could use ES202
Data (QCEW) but not readily
available--in this case
used data provided
by the three auto
companies
Where have all the auto jobs gone?
Plant Location
~2007-2008
11
Plant Location
~2007-2008
Where have all the auto jobs gone?
12
Where have all the auto jobs gone?
Source: Whole Data
13
Michigan Employment Dynamics
20
18
16
14
12
10
8
6
4
2
0
3361
18.06
3363
16 27
16.27
12.12
9.31
8.05
4.11
0.28
5.65
4.35 4.58
1 17
1.17
new hires
0.98
separations
2001
2005
new hires
separations
2008
Source: QWI
Where have all the auto jobs gone?
14
Where have all the Auto Workers Gone?
Michigan Motor Vehicle Manufacturing
Separations by Quarter
14,000
12,000
Under age 25
25-34
35-44
45-54
55-64
65 and older
10,000
8,000
6,000
4,000
2,000
0
2005
2006
2007
2008
By far, the most separations have occurred among workers age 55 to 64. These
individuals are near retirement and may not try to re-enter the workforce.
Source: MDLEG, local employment dynamics QWI data.
15
Auto Worker Transition from Filing UI Claim to
Re-employment,
Re
employment, 2006
Recall
27283
(50 2%)
(50.2%)
UI Claim
54328
(13.4%)
No recall
27045
(49.8%)
Not Employed
33 0
3370
(12.5%)
WIA
646
(19.2%)
Employed yr 1
in same ind.
(73.7%)
Employed
(87.5%)
Employed yr 2
in same ind.
(57.3%)
Employed
(78.1%)
Employed yr 1
in different ind.
(26.3%)
Employed yr 2
in same ind.
(42.7%)
Employed yr 1
Employed yr 2
in same ind.
in same ind.
(27.0%)
Employed
Employed (24.6%)
(79.6%)
(77.7%)
Employed yr 2
Employed yr 1
In different ind.
in different ind
ind.
(63.0%)
(75.4%)
16
Where have all the auto workers gone?
Non-Auto Worker Transition from Filing UI Claim to
Re-employment,
Re
employment, 2006
Recall
56040
(15 4%)
(15.4%)
UI Claim
363911
No recall
307871
(84.6%)
Not Employed
4936
49367
(16.0%)
WIA
6572
(13.3%)
Employed yr 1
in same ind.
(56.9%)
Employed
(84.0%)
Employed yr 2
in same ind.
(47.2%)
Employed
(77.7%)
Employed yr 1
in different ind.
(43.1%)
Employed yr 2
in same ind.
(52.8%)
Employed yr 1
Employed yr 2
in same ind.
in same ind.
(16.8%)
Employed
Employed (14.9%)
(80.5%)
(81.4%)
Employed yr 2
Employed yr 1
In different ind.
in different ind
ind.
(83.2%)
(85.1%)
17
Where have all the auto workers gone?
Quarterly Earnings of UI Applicants Not Exempt from Job
Search
10,000
9,000
8 000
8,000
7,000
6,000
Auto Workers
5,000
Non-Auto Workers
4,000
3,000
2,000
1,000
16
14
12
10
8
6
4
2
-2
-4
-6
-8
-1
0
-1
2
0
Quarters Before and After UI Application
Where have all the auto workers gone?
18
Earnings of Incumbent Workers and New Hires
in Michigan’s Auto Industry
$7,000.00
89%
90%
$6,000.00
68%
76%
$5,000.00
56%
56%
$4,000.00
Incumbent
New Hires
$3,000.00
$2,000.00
$1,000.00
$0.00
3361
3363
All Ind.
2004Q4 2005Q3
2004Q4-2005Q3
Where have all the auto workers gone?
3361
3363
All Ind.
2007Q4
2007Q4-2008Q3
2008Q3
Source: QWI
19
Summary
•
•
•
•
•
•
The restructuring of the auto industry, particularly in the traditional
manufacturing regions, has caused significant worker displacement
and community upheaval
Proper responses require informed decision-making
Requires answers to the questions posed here and to more in-depth
questions specific to each region
Great strides have been made in providing useful data to states and
local areas
We have access to better data (e.g., flows) and more analytical
capabilities than ever before
But I believe we are only scratching the surface in how these data
can be used to tackle such problems
20
Summary
•
•
•
•
Besides making UI, education, business, and administrative data
more available for analysis, we also need to think about how we can
put this information at the finger tips of stakeholders (front-line staff)
to make more informed decisions on a more real-time basis
We also need to figure out how to canvas the needs of businesses
more effectively and better assess and map the capital assets of a
region
The current economic crisis provides an opportunity to step up and
provide the information that will provide a more effective,
transparent, and effective delivery of services
And create a culture of evidence-based decision making
21