Transcript Organizations in America: analyzing their structures and human

SOCI 5013: Advanced Social
Research
The 1991 National Organizations
Survey (NOS)
Spring 2004
part of this set of lecture notes comes
from presentation by Cynthia Kendall,
Arthur Heether and Kohei Takahashi
A
NOS 1991
• For decades, sociologists have been
studying workplace structures and resources
allocations.
• However, available empirical data from
U.S. census bureau provide little knowledge
of workplace information.
Existing Dataset
• Large national datasets such as CPS, SIPP,
PSID all have workplace module, which
collected information from workers’
perspective.
• Lack of available dataset restricts
organization researchers to using qualitative
methods such as case studies, in-depth
interviews to investigate a very limited
number of organizations.
NOS 1991
• The situation drastically changed when the
1991 National Organizations Survey (NOS)
comes out in 1990s.
• Indeed, the NOS 1991 is the most
successful national organizations survey in
American Sociology during the 20th
century!!
Principal Investigators (PI)
•
•
•
•
Arne Kalleberg (Univ. of North Carolina)
David Knoke (Univ. of Minnesota)
Peter Marsden (Harvard University)
Joe Spaeth (NORC, Univ. of Chicago)
The Success
• Publications from the 1991 GSS/NOS include at
least 20 manuscripts in peer reviewed journals,
several book chapters, a book.
• Kalleberg, Knoke, Marsden, and Spaeth. 1996.
Organizations in America: analyzing their structures
and human resource practices Thousand Oaks,
Calif.: Sage Publications
• The emergence of 1991 NOS changed the landscape
of organization studies for Sociologists and Business
Professors.
Goals of NOS
• To understand American organizations
structuration such as bureaucracy and high
performance organizations
• Better understand complex organizations,
and social stratification within organizations
setting
Goals of NOS
• Understand better whether American
organizations are operated as a atomic close
systems, or open systems,
• How the contextual factors in technical and
non-technical aspects effect organizations
in-practice
Goals of NOS
• Internal organizations structure and its
pioneer researchers are Blau and
Schoenherr (1971) about formalization, and
Aston group 1969, who found tremendous
variation among orgs in their differentiation,
formalization, and decision centralization
Goals of NOS
• External technical influence with its pioneer
researcher of Lawrence and Lorsch 1967,
who found complexity and uncertainty are
key dimensions to which organizations must
adjust
Institutionalization
• Institutional theorists (Meyer and Rowan
1977; DiMaggio and Powell 1983) stated
that non-technical aspects of
institutionalization exerts external effects on
organizational in-practice.
Institutional Theory
• Central concerns of old and new
structualism is that individual socialeconomic outcomes not only result from
individual attributes, but also consequences
of supraindividual constructs such as
classes, labor market sectors.
Matched Dataset
• A critical questions is how to collect data
from establishment (organization) and
individuals working in the establishments
levels.
• 1990 Lincoln and Kalleberg surveys of
manufacturing plants and their employees in
Indianapolis and Atsugi, Japan
Matched Dataset
• Bridges and Villemez 1994 studies modern
HR practices in Chicago, Illinois area.
• Tomaskovic-Devey 1993 studies of gender
and racial segregation in establishments and
workers in North Carolina.
Methodology
• Researchers encountered many difficulties
and made great accomplishments in
implementing the NOS 1991 design, which
is detailed documented in Spaeth and
O’Rourke (1994).
Methodology
• Many theoretical concerns in previous section
requires cross-level data of both individuals
and their workplaces.
• Several levels of organization survey: work
position, subunits, establishments/workplaces,
firms
Which one to Use
• Establishment, defined as a workplace at a
specific geographic location.
• UARK credit union at the 6th street is an
establishment, whereas UARK credit union
is a firm
Work Establishment
• An establishment/workplace encompasses
most employment, and managers are quite
knowledgeable about their HR policies.
• Some issues are pertinent to individuals, so
the research design also encompasses
questionnaire items for individual workers
within establishments
Sampling
• Organization surveys have generalizability
issue
• Survey with fortune 500 gravitates toward
large size establishments,. Findings from
those survey cannot be readily generalized
to represent many median and small size of
American workplaces
Generalizability
• Findings from surveys within a certain
industry such as informational technology,
pharmaceutical industry cannot be
generalized to other types of industries. But
the trade off is that those surveys generally
yield detailed background info.
National Representatives
• Many countries have a comprehensive
rosters of employers, such as England,
Norway, and Sweden. U.S. does not (the
new Dun’s list have several major problems
but was later improved).
• Multiplicity and hyper-network sampling, in
essence, this design ensures that random
sampling of individuals necessarily yields a
random samples of employers.
NOS sample
• Because large organizations are more likely
to be selected due to this random sampling
from individuals, this method equals PPS.
• Consequently, NOS 1991 has a
disproportional large size organizations.
Cross-level
• NOS 1991 is linked to GSS 1991. ICPSR
contains both NOS and GSS 1991, one
needs to download both dataset and
carefully match up each case to form a
combined dataset of NOS/GSS dataset,
which is used in Yang’s (2003) paper
examining structural (establishments)
interactions with occupations (individuals)
in affecting commitment (individuals).
Implementations
• 1991 General Social Survey (GSS) done by
NORC
• Individual survey produces information for
1,127 workplaces
• Survey Research Laboratory (SRL) at
University of Illinois at Urbana-Champaign
completed 727 workplace interviews
Informants
• Head of personnel department
• Functional equivalent personnel
ACTIVITY
N
%
N
%
______________________________________________________________________________
TOTAL OCCUPATIONAL INFORMATION
INITIALLY INADEQUATE INFORMATION
1, 427
100. 0%
371
26. 0%
GSS Respondent Refusal
Usable
Unusable
172
6
166
100. 0%
3. 5%
96. 5%
Missing Data on Name/Address
Usable
Ineligible
Unusable
199
125
5
69
100. 0%
62. 8%
2. 5%
34. 7%
1, 056
996
54
6
100. 0%
94. 8%
3. 8%
0. 6%
INITIALLY ADEQUATE INFORMATION
1, 056
74. 0%
Initially Adequate Information
Usable
Ineligible
Unusable
TOTAL OUTCOMES
Usable
Ineligible
Unusable
1, 427
1, 127
59
241
100. 0%
79. 0%
4. 1%
16. 9%
Results of SRL Interview Attempts
______________________________________________________________________________
ACTIVITY
N
%
N
%
______________________________________________________________________________
TOTAL USABLE INFORMATION
DUPLICATE EMPLOYERS
1, 127
100. 0%
60
5. 3%
All Duplicates
Refused
Pending
Completed
60
14
7
39
100. 0%
23. 3%
11. 7%
65. 0%
TELEPHONE INTERVIEWS ONLY
Refused, No Mail
Pending, No Mail
Completed by Telephone
633
54
18
561
100. 0%
5. 1%
1. 7%
52. 6%
MAIL QUESTIONNAIRE
Refused
Pending
Completed Mail Questionnaire
434
192
115
127
100. 0%
44. 2%
26. 5%
29. 3%
SRL INTERVIEW ATTEMPTS
TOTAL OUTCOMES
Refusal
Pending
Completed
1, 067
1, 127
260
140
727
94. 7%
100. 0%
23. 1%
12. 4%
64. 5%
Questionnaire Design
• Core occupation: the job title for workers
directly involved in core production or
service delivery of work establishments
• Core occupation for University of Arkansas
• GSS occupation
• Individual level data
• Workplace level data
Publications from the 1991 NOS
• Organizational structure (Marsden et al., 1994),
which investigates market scope, sector and
union presence structural types
• Absence of formalization (Bothner 1998)
• High Performance Work Systems (HPWS)
Characteristics (Kalleberg and Moody 1996)
Publications
• Large, manufacturing, unionized, multi-site
workplaces and “post-Fordist” trends (Handel
1998)
• Procedural due process (Kallenberg et al., 1996)
• Public sector organizations (Sutton et al.,1994)
Publications
• Managerial gender integration (Reskin
and McBrier 2000)
• Job search process: difference in reliance on
referrals (Marsden and Gorman 1998)
• Job training (Knoke and Kalleberg 1994)
Publications
• Income and sex composition (Huffman and
Velasco 1997). Main finding is 1% increase in
percentage of women on job is associated with
$180 reduction in annual earnings for core and
managerial/administrative jobs
• Non-monetary employee benefits (Knoke 1996)
Publications
• Kallenberg and Van Buren (1996) used a
combined dataset, the 1991 GSS/NOS to reinvestigate work benefits including medical
and dental coverage, life insurance, sick
leave, maternity leave and pension plans
Publications
• Public sector and firm/internal labor
market–Likelihood of promotion (Kalleberg
and Van Buren 1996)
• Sex differences in promotions (Kalleberg
and Reskin 1995)
• Other intrinsic rewards such as commitment
to organizations (Kalleberg and
Mastekaasa 1994, Yang 2003)
Discussions
• What objectives were the NOS research
team unable to achieve and how did they
restructure their design ?