Abt
Associates
of Canada

Social
Research
Consultants

Project Report

MINORITY TARGET GROUPS:
AN ANALYSIS OF DATA FROM
THE SURVEY OF BARRIERS
TO ADVANCEMENT IN THE
PUBLIC SERVICE

Prepared for

Greg Hunter
Human Resources Division
Treasury Board of Canada

Participating Staff

David Hoffman
Frank Eaton

January 1990
Ottawa
N3482/ed

I

INTRODUCTION

A.

BACKGROUND
In September 1988, the President of the Treasury Board established a Task Force

on Barriers to Women in the Public Service. The Task Force engaged Statistics Canada to

conduct a survey of public servants to provide information on the barriers to advancement
they had encountered and on the actions public servants thought would help most in over­

coming possible barriers. Consultants assisted Statistics Canada in the development of the
questionnaire by conducting focus groups across the country. In May 1989, Statistics

Canada distributed 20,000 questionnaires to a sample of public servants. The frame for the

sample was the incumbent file of December 1988, filtered to include only employees
working under the PSSRA in Canada and to exclude employees in term positions of less

than 6 months. Of these, 12,044 public servants returned questionnaires, making a
response rate of 60%.
The objective of the Task Force survey was to determine whether or not women in

the federal public service face barriers to advancement and whether these barriers differ

from barriers facing men. The data are rich in information about the experience of public
servants with respect to working conditions, barriers to access to developmental opportuni­

ties and promotion and other job-related matters. While the original purpose of the survey
was to permit exploration of gender-based differences among public servants, responses to

the survey can be usefully analyzed to explore a variety of other concerns as well.

Treasury Board asked Abt Associates to re-analyze the survey results focussing on

barriers for the disabled, aboriginals and visible minority employees of the Public Service.

Abt Associates of Canada

1

Our task has been to focus the analysis and the report on situations where the
experiences and perceptions of the above-mentioned equity target groups are significantly

different from those of other public servants. We have directed our analytical efforts at
exploring four main questions:

B.

>

Do target group members differ significantly from other public servants in
their rate of advancement?

>

Have target groups who have not changed position since entry into the
public service spent a shorter or longer period of time in their current posi­
tion than comparable other public servants?

>

Does being a member of a target group affect denial of promotion or
development opportunities and perceptions of fairness of promotion in the
public service?

>

Do target group members face different barriers from other public servants?
And to what extent do they mention barriers specifically related to being
handicapped, aboriginal or a member of a visible minority?

REPORTING "SIGNIFICANT" RESULTS
In this report when we call attention to differences between groups, we do so when

they are statistically significant. When a result is reported as statistically significant,

we mean that the result, though based on a sample, would likely be true if all public
servants were surveyed. Or, to state the converse, two differences of proportion that are

said to be ’’not statistically significant” may look noteworthy, but they are based on sample
sizes that are too small to be sure that the result would be verified if everyone in the groups

was surveyed.
Much of our analysis relies on chi-square tests of statistical significance. This test
is appropriate for comparing proportions of public servants who possess a certain

characteristic, such as having sought a promotion in the past three years.
To increase our confidence in the reporting of the findings of the survey we have

conducted a considerable number of analyses using multivariate statistical techniques such
as multiple linear regression and discriminant analysis. The purpose of these technically

Abt Associates of Canada

2

complex procedures is to try to ensure that statistically significant relationships between
demographic and job-related variables and experience or perceptions are not in fact better
explained by combinations of these and other factors that we might have otherwise have
overlooked.

C.

ANALYTICAL STRUCTURE

Our analysis of the questions listed above follows a common format for the three

target groups under study. Therefore, we briefly describe our analytical approach to each
question here, then present the results for each group separately in the next chapter.
1.

Rate of Advancement
We seek to test the hypothesis that members of a target group advance more

slowly than others in the public service. Our analysis also examines how target group

members who have advanced rapidly differ from those who have not.
As in earlier work, we use change in average (mean) salary as a proxy for

career advancement. Data added to the survey file by Statistics Canada included the mean
salary in the employee's current group and level (MSC) and that of the group and level in

which he or she entered the public service (MSE), adjusted for inflation. Comparing the

two provides a measure of an individual’s career advancement
We recognize that a valid measure of career advancement is difficult to
define and to construct Advancement or success can comprise many forms of achievement

or reward that may not be adequately reflected in the measure used in our analysis.
However, within the constraints imposed by the data and by the underlying assumptions of
our analysis, the above measure best served our analytical purposes.
We point out that the measure does not directly represent a rate of advance­

ment As we explain below, however, we control for years of service in the models we

use. Therefore, while we do not examine rates of advancement directly, we achieve the

Abt Associates of Canada

3

same result by comparing amount of advancement among people with comparable duration
of service. When we state (below) that one group advances "more quickly" or "at a greater
rate" than another, we.are in fact referring to "greater advancement over the same number

of years of service."
For public servants who have advanced, we use the measure MSC-MSE.

To satisfy assumptions underlying our analytical methods, we work with the natural
logarithm of this difference. Hence the measure of advancement we analyze is log(MSC-

MSE), where MSC-MSE is greater than zero.
Our comparison of advancement of target group members and non-members

uses two multiple regression models in which advancement (as defined above) is the
dependent variable. The independent (explanatory) variables used are listed in Appendix

A. A test of statistical significance on the estimated1 coefficient for a variable indicating
target group membership determines whether membership has a significant effect on
advancement, overall. In the second model, we allow for variations in this effect within
classes of Category and Level combinations. (Statistics Canada has defined generic Junior,

Intermediate, and Senior levels across standard classification Groups for this purpose.)

The exact combinations used varies among target groups, depending on the number of
respondents available in each. The interpretation of the results of these models is less

straightforward, and we refer technically inclined readers to Appendix A for details.
Applying the first model to only target group members allows us to identify
characteristics associated (in a statistically significant way) with lesser or greater advance­

ment scores among members of a specific target group.

A statistically significant

coefficient on any variable in the model indicates its influence.
2.

Index of Stagnation

Many respondents had not changed group and level since entering the public
service. Put another way, they experienced zero advancement (MSC-MSE=O). Since non­

advancers likely differ considerably from advancers in many ways, we study the two
groups separately. Years of service (LGNS) is the analytical variable we use for people

with zero advancement. It embodies the converse concept of lack of advancement,

Abt Associates of Canada

4

indicating the duration of the state of having failed to advance, and may be considered an

index of stagnation in career advancement

Length of service is the dependent variable for a second set of regression
models used to analyze the index just described. The models follow much the same
structure as those described above for the analysis of advancement scores. They exclude

certain variables that distinguish conditions of the first public service job from those of the
current one, since both are the same for employees who have not advanced. Again,

Appendix A shows the exact set of variables used.
3.

Denial of Promotion or Development and Perceptions of
Fairness
This component of our analysis involves six variables derived from several

questions from the original survey. If the respondent indicated in Question 24 that he or
she wanted or requested a promotion in the last three years. Question 25 addressed
perceived reasons why the respondent might have missed out on, or been denied, a promo­

tion. We interpret any response to Question 25 as an indication of denial of promotion, and

this forms our first analysis variable. Similarly, if the respondent indicated in Question 27

that he or she wanted or requested a developmental opportunity (secondment, training, etc.)
in the last three years, Question 28 addressed perceived reasons why the respondent might

have missed out on, or been denied, such an opportunity. Again, we treat any response to

Question 28 as an indication of denial of a development opportunity, which is the second
analysis variable.
Question 35 asked respondents to indicate to what extent they agree with

several perceptions. We have selected four perceptions that have particular bearing on
advancement of target groups:
(2)

Positions are often posted after the department/agency has already
identified the person they want to fill the position.

(9)

Employment equity and affirmative action programmes give women
an unfair career advantage when the public service is downsizing.

Abt Associates of Canada

5

(15)

People get ahead because of who [sic] they know.

(16)

People quit the public service rather than repay pension contribu. tions after extended leave.

A response indicating either slight or strong agreement with each of these perceptions was

the basis for four more analysis variables.
The method of analysis for determining the effect of membership in a target

group is the same for all six variables. Each is a binary (dichotomous, or "dummy")
variable, taking the value 1 if a condition is present and 0 otherwise. Therefore, we

selected discriminant analysis as the method for detecting an effect due to membership in a
target group. This method is commonly applied in cases where the outcome variable has

this structure. The discriminating variables considered for the analysis were essentially
those that entered the regression models described above. We used a stepwise approach
designed to select one such variable at a time. The variable selected at each step was the

one with the greatest discriminating power among all the remaining candidates. Details of

the results of these analyses are presented in Appendix A.
4.

Comparability of Barriers Faced

The focus of this analysis is a comparison of barriers faced by members of
target groups, compared to those faced by non-members.

The barriers are those originally listed in Question 25 of the questionnaire (1

to 29) as well as a further 6 responses that were developed after the survey was completed
when "other, please specify" was broken out further. The proportion of respondents

answering Question 25 is reduced below the total numbers in each target group; those who

were not denied promotion did not answer. On the other hand, multiple responses were
possible. The analysis is based on cross tabulations and our interest focuses on those

situations in which the target group members differ significantly in the barriers they cite
from other public servants.

Abt Associates of Canada

6

D.

THE TARGET GROUPS
Exhibit 1-1 shows the distribution of the three target groups by occupational

category and level (junior, intermediate and senior). The numbers shown should be inter­

preted as follows:
>

The column headed "N” is the unweighted number of cases in the sample.

>

The column headed ’’Population” is the weighted number, representing the
number in the particular level and category of the public service as a whole.

The exhibit shows that there are 209 aboriginal respondents representing 3,487

aboriginals in the public service population as whole; 667 disabled representing 13,216 in
the population; and 516 visible minority members, representing 8,985 in the population.
Question 44, the source of data on target status was constructed such that respon­

dents could indicate multiple target group memberships. The vast majority of target group

members indicated only a single group membership. However, the following combina­
tions of target group membership were reported.
>

Disabled and visible minority only ~ 41.

>

Disabled and aboriginal only - 15.

>

Visible minority and aboriginal only — 21.

>

Disabled, aboriginal and visible minority -- 5.

These combined target groups are too small to warrant separate analyses.

Abt Associates of Canada

7

EXHIBIT 1-1

Number of Target Group Respondents and Number in the
Public Service Population - By Occupational Category
and Level

N

Popu­
lation

N

Popu­
lation

Visible
Minority
Popu­
N
lation

Aboriginal
Category

Level

Disabled

Management

Senior

5

22

11

377

14

85

Scientific &
Professional

Junior
Intermediate
Senior

5
8
7

66
30
16

28
21
11

377
366
27

50
90
26

649
1,608
69

Administrative &
Foreign Service

Junior
Intermediate
Senior

38
19
11

704
280
' 55

91
46
22

1,959
873
149

80
34
13

1,716
579
67

Technical

Junior
Intermediate
Senior

15
3
-

310. 61
13 • 27
19:

1,520
341
34

31
14
5

562
161
9

Administrative
Support

Junior
Intermediate
Senior

35
25
-

811
404
-

109
95
2

2,435
1,602
3

71
42
2

1,630
709
3

Operational

Junior
Intermediate
Senior

17
18
4

328
424
25

38
68
18

1,123
2,235
110

18
24
2

452
675
10

209

3,487

667

13,216

516

8,985

ALL

Abt Associates of Canada

8

II

FINDINGS WITH RESPECT TO THE THREE TARGET GROUPS

A.

ABORIGINALS

The analysis in this section concerns survey respondents who identified themselves

(Question 44) as Inuit, Metis, Non-Status Indian, or Status Indian. For convenience, we
refer to this group as aboriginals.
1.

Advancement (for those who advanced)

Overall, aboriginals experienced no significantly different advancement than
non-aboriginals. Although the regression estimate indicated about 8.2% less advancement

for aboriginals over comparable length of service, the data were not precise enough to give

us confidence that the difference occurred due to reasons other than sampling error.
Examination of the model disaggregated by classes of category and level

also revealed no significant differences between aboriginals and non-aboriginals.

Looking only at aboriginals, we found that several factors have significant

influence on advancement Associated with greater advancement are the following
characteristics:
>

Female

>

Longer period of service (years since first PS job)

>

First PS job was a term position

►

Longer period(s) of leave taken for educational purposes

>

Current position in intermediate or senior levels of Administrative
and Foreign Service Category (relative to junior Administrative and
Foreign Service, which we used as the reference group)

Abt Associates of Canada

9

Employees with the following characteristics exhibited less advancement:

2.

>

Older

>

More children under five years old

>

Longer period(s) of leave taken for pregnancy

>

Current position in Technical, Administrative Support, or Opera­
tional category (relative to junior Administrative and Foreign
Service)

Stagnation (for those who did not advance)
Again, among employees who did not advance to a group and level with a

higher mean salary, aboriginals had not stayed in.their positions any longer than non­

aboriginals.
In the disaggregated analysis, however, we found that aboriginals in the

Technical category had remained in the same group and level 6.3 years longer than
comparable non-aboriginals. The survey received only 18 responses from aboriginals in

this category, but the large difference in years of service achieved statistical significance.

Among aboriginals, the following statistically significant factors increased
the measure of stagnation:

3.

>

More children between 12 and 17 years old

>

Educated to the Master’s level (compared to failing to finish high
school, which was the reference group)

>

Current position in the Technical category (compared to junior
Administrative and Foreign Service)

Likelihood of Identifying Specific Barriers
This section relies on discriminant analysis to identify the factors associated

with an individual's perception of the existence of certain unfavourable characteristics of

Abt Associates of Canada

10

promotion opportunities in the public service or of having been denied a promotion or a
developmental opportunity.

The characteristics about which the survey solicited opinions and which we

examined in this study were described in Chapter I. However, whether a respondent was
aboriginal or not did not significantly affect agreement with any of the cited statements or
denial of a promotion or a developmental opportunity, after controlling for other

influencing factors measured in the survey.
4.

Differences in Barriers Identified

Exhibit II-1 compares the barriers identified by aboriginals with those

reported by other public servants. The percentages reported are for public servants who

reported being denied promotions. Although 15% of aboriginals denied promotion said
they were denied promotion because of their aboriginal status, this reason for denial was

not the most frequently cited reason. More important for aboriginal respondents were the
following reasons:

>

I was not part of the group - 30%.

>

Manager did not think I was ready - 28%.

>

Manager did not want to to replace me - 27%.

>

Manager would not support my application - 25%.

>

Manager did not give me information -21%.

Note that these reasons were given in roughly the same proportions as by other public
servants. However, also noteworthy is the fact that 8% said they were denied promotion
because they objected to sexual harassment and 15% said that it was because they were

women.

Abt Associate* of Canada

11

EXHIBIT II-l

Barriers Identified By Aboriginals

Aboriginal

ALL

Yes

No

Unweighted N|Weighted weighted x|Unweighted N weighted ^Weighted % unweighted N|Weighted ^Weighted X

TOTAL DENIED
Manager did not think I was ready
Manager did not want to replace me
Manager did not give me information
Manager would not support my application
Did not get along with supervisor
Did not want overtime, set hours, shifts
I was too young
I was too old
I was or was planning to become pregnant
I am a woman
I am a man
I am a member of a visible minority
I am a person with disabilities
I am an aboriginal person
Would not normally hire a woman
Would not normally hire a man
I had young children, dependents at home
I worked part-time or shared a Job
I was divorced, separated
I was a single parent
I objected to sexual harassment
Took matemity/patemity/parental leave
Took language/education leave
Took long term dlsabllity/sick leave
I am an anglophone
I am a Francophone
I could not relocate
I was not "part of the group"
I was unable to travel
I did not meet a language requirement
Downsizing, budget cut
Competition, promotion process not fair
Nothing available
I lack educational requirement
Other

4,034
834
971
558
863
451
68
228
210
93
417
296
137
66
1
131
30
132
62
28
31
43
91
20
55
801
206
238
1,184
52
51
49
118
305
53
1,020

68,799
14,074
16,945
10,441
15,841
8,041
1,229
3,715
3,542
1,313
5,181
5,653
2,582
1,268
21
1,446
587
2,080
960
499
505
638
1,414
313
1,162
13,089
3,702
3,674
20,552
859
714
771
2,271
5,326
1,013
17,581

100.00
20.46
24.63
15.18
23.02
11.69
1.79
5.40
5.15
1.91
7.53
8.22
3.75
1.84
0.03
2.10
0.85
3.02
1.40
0.73
0.73
0.93
2.05
0.46
1.69
19.02
5.38
5.34
29.87
1.25
1.04
1.12
3.30
7.74
1.47
25.55

70
19
18
16
15
8
0
4
3
0
16
2
7
0
10
6
2
3
0
1
1
6
1
1
0
11
1
' 5
22
1
0
1
1
3
1
21

1,160
329
310
243
292
113

100.00
28.39
26.74
20.91
25.20
9.77

78
22

6.72
1.87

170
42
141

14.66
3.59
12.19

179
51
33
44

15.47
4.43
2.82
3.77

10
25
97
15
5

0.88
2.15
8.35
1.31
0.43

136
25
90
350
15

11.72
2.14
7.80
30.15
1.31

6
2
57
25
352

0.53
0.16
4.91
2.15
30.32

4,104
853
989
574
878
459
68
232
213
93
433
298
144
66
11
137
32
135
62
29
32
49
92
21
55
812
207
243
1,206
53
51
50
119
308
54
1,041

69,959
14,403
17,255
10,684
16,133
8,154
1,229.
3,793
3,563
1,313
5,351
5,694
2,724
1,268
200
1,497
620
2,123
960
510
530
735
1,429
318
1,162
13,225
3,727
3,764
20,902
874
714
777
2,273
5,383
1,038
17,933

100.00
20.59
24.66
15.27
23.06
11.66
1.76
5.42
5.09
1.88
7.65
8.14
3.89
1.81
0.29
2.14
0.89
3.04
1.37
0.73
0.76
1.05
2.04
0.45
1.66
18.90
5.33
5.38
29.88
1.25
1.02
1.11
3.25
7.69
1.48
25.63

B.

DISABLED

This section concerns survey respondents who indicated (Question 44) that they

consider themselves, or believe that a potential employer would likely consider them,
disadvantaged by reason of a persistent and severe disability. For short, we refer to this
group as disabled.

1.

Advancement (for those who advanced)
The disabled experienced advancement no different from that of the able-

bodied. The regression estimate indicated about 2.2% greater advancement for the
disabled, but this amount was not statistically significant at even close to the.05 level.

Disaggregating by category and level reinforced this finding. While some

differences (see Appendix A) were quite large, inherent variation in the data prevented a
conclusion that the differences were due to other than sampling error.
Examining the disabled only, we found several factors that influence

advancement in a statistically significantly fashion. Associated with greater advancement
are:

>

Completion of high school (compared to not completing)

>

Longer period of service (years since first PS job)

>

First PS job was in the National Capital Region

>

First PS job was a term position

>

Employee was denied a developmental opportunity in the past two
years

>

Current position in the Managerial category or at the intermediate or
senior levels in one of the Scientific and Professional, Administra­
tive and Foreign Service, or Technical categories (compared to
junior Administrative and Foreign Service)

Abt Associates of Canada

13

The following characteristics tend to curtail advancement among the

disabled:

2.

►

Older

>

Current position at the junior level in the Technical category, or at
any level in the Administrative Support or Operational categories
(compared to junior Administrative and Foreign Service)

Stagnation (for those who did not advance)

We find some evidence that disabled employees who did not advance tended
to have stayed at the same group and level slightly longer than the able-bodied. The

regression model yielded an estimate of two-thirds of a year longer for this difference.

While this value is not significant at the.05 level, it.would be at the.08 level. We conclude
that the disabled may stagnate longer than the able-bodied, but warn that this conclusion

has a greater chance of being incorrect than others we state in this report (8% versus 5% or
less).
This effect was particularly evident at the intermediate and senior levels of

the Operational category. The regression analysis indicates that disabled people who have

been in such positions since entering the public service (86 respondents fit this description)
have been there 2.9 years longer than comparable able-bodied employees.
Looking at only the disabled who failed to advance, we find that older

workers have stayed longer at the current group and level, while those for whom the first

public service position was a term position had a shorter stay there, as did people at the
intermediate or senior levels in the Scientific and Professional category or at the junior level

in either the Administrative Support or the Operational categories (compared to junior
Administrative and Foreign Service).
3.

Likelihood of Identifying Certain Barriers
We found that being both disabled and at the intermediate or senior levels of

the Operational category was significantly associated with agreement with the statement that

Abt Associate* of Canada

14

people will quit the public service rather than repay pension contributions after an absence.
Among people in this group, 18% agreed, compared to 10% of those not in this group.
In no other cases (denial or promotion or developmental opportunity or
agreement with the other three statements studied) did being disabled have any significant

influence on the respondents' answers.
4.

Differences in Barriers Identified
Exhibit II-2 shows the pattern of responses for the disabled, compared to

other public servants denied promotion.

As was the case with aboriginal respondents, being a member of the target

group was not the most frequently cited reason for denial of promotion. Not being part of
the group (38%), not having the manager's support for an application (29%), having a
manager unwilling to replace (26%) or who thought them unready (24%), were all more

frequently cited that "I am a person with disabilities'* (17%). Overall, apart from the
obvious difference with respect to identifying being a person with a disability as a reason

for not being promoted, and not being part of the group, the profile of reasons given the
disabled is very similar to that given by other public servants.

C.

VISIBLE MINORITIES

The analysis in this section concerns survey respondents who identified themselves
(Question 44), by virtue of race or colour, as a member of a visible minority.

1.

Advancement (for those who advanced)

Members of visible minorities have experienced greater advancement than
non-members. Overall, respondents who identified themselves as members of a visible

minority had an advancement score 8.8% higher than comparable non-members.

Abt Associates of Canada

15

EXHIBIT II-2

Barriers Identified By Disabled

■ Disabled
No

ALL

Unweighted N|Weighted n Weighted x|Unweighted N|Weighted N|Weighted %|Unweighted n|Weighted N Weighted X
_
4——
100.00
100.00
4,104
69,959
5,149
255
100.00
64,810
3,849
TOTAL DENIED
20.59
24.43
853
14,403
1,258
20.28
58
13,145
795
Manager did not think I was ready
24.66
17,255
26.35
989
67
1,357
24.53
15,898
922
Manager did not want to replace me
15.27
17.67
574
10,684
910
44
15.08
9,774
530
Manager did not give me Information
23.06
28.95
878
16,133
67
1,491
22.59
14,642
811
Manager would not support my application
11.66
8,154
47
911
17.69
459
11.18
7,243
412
Did not get along with supervisor
1.76
1,229
2.37
68
6
122
1.71
1,107
62
Did not want overtime, set hours, shifts
5.42
3,793
6.46
332
232
15
5.34
3,461
217
I was too young
5.09
3,563
305
5.93
213
16
5.03
3,258
197
I was too old
1.88
1.73
93
1,313
89
1.89
5
1,224
88
I was or was planning to become pregnant
7.65
7.48
433
5,351
385
23
7.66
4,966
410
I am a woman
8.14
5,694
7.58
298
390
18
8.18
5,304
280
I am a man
3.89
2,724
144
395
7.67
18
3.59
2,329
126
I am a member of a visible minority
1.81
1,268
887
17.23
66
44
0.59
380
22
I am a person with disabilities
0.29
0.41
200
11
21
0.28
1
179
10
I am an aboriginal person
2.14
1,497
3.16
137
163
12
2.06
1,334
125
Mould not normally hire a woman
0.89
620
1.39
32
71
3
0.85
549
29
Mould not nonsally hire a man
3.04
2,123
2.47
135
127
3.08
8
1,996
127
I had young children, dependents at home
1.37
960
0.48
62
25
1.44
1
936
61
[ worked part-time or shared a Job
0.73
510
0.88
29
45
2
0.72
464
27
I was divorced, separated
0.76
530
0.33
32
17
1
0.79
513
31
I was a single parent
1.05
735
1.24
49
64
1.04
3
671
46
I objected to sexual harassment
2.04
1,429
2.22
92
114
6
2.03
1,315
86
Took maternity/paternity/parental leave
0.45
318
21
16
0.30
0.47
1.
303
20
Took language/education leave
1.66
6.48
55
1,162
334
13
1.28
828
42
Took long term disability/sick leave
18.90
13,225
21.23
812
1,093
54
18.72
12,132
758
I am an anglophone
5.33
207
3,727
5.06
261
12
5.35
3,467
195
I am a Francophone
5.38
3,764
243
296
5.75
• 15
5.35
3,468
228
I could not relocate
29.88
1,206
20,902
1,947
37.81
94
29.25
18,955
1,112
I was not "part of the group"
1.25
874
1.99
53
102
4
1.19
772
49
I was unable to travel
1.02
714
64
1.24
51
1.00
5
651
46
I did not meet a language requirement
777
1.11
50
0
1.20
777
50
Downsizing, budget cut
3.25
2,273
119
3.43
176
10
3.23
2,097
109
Competition, promotion process not fair
7.69
5,383
308
6.68
344
15
7.78
5,039
293
Nothing available
1.48
1,038
54
1.73
89
1.46
3
949
51
I lack educational requirement
25.63
1,041
17,933
27.26
1,404
69
25.50
16,529
972
Other

At the disaggregated level (by category and level), the analysis does not
reveal any significant differences. We conclude that the effect overall achieves statistical
significance due to the larger numbers of members of visible minorities contributing to the

single effect in that model, while the numbers are not sufficient at the disaggregated level to

do so.

Considering only respondents who are members of visible minorities, we
find that the following factors contribute to greater advancement:
>

Longer period of service (years since first PS job)

►

Current position is in the Managerial category or at the intermediate
level in the Scientific and Professional category (compared to junior
Administrative and Foreign Service)

At the same time, advancement is slower for members of visible minorities

who possess the following traits:

2.

>

Older

>

Current position at the junior level in the Scientific and Professional
category, at the intermediate or senior level in the Administrative and
Foreign Service category, or in the Administrative Support or
Operational categories (compared to junior Administrative and
Foreign Service)

Stagnation (for those who did not advance)
Members of visible minorities who have not advanced do not languish in

their current positions longer than comparable counterparts. Both overall and disaggre­
gated by category and level, there was no statistically significant difference in the length of

time spent at the current group and level.
Among members of visible minorities, age was the only factor that was
associated with length of time at the same group and level. Older workers tended to have

remained longer.

Abt Associates of Canada

17

3.

Likelihood of Identifying Specific Barriers

Being a member of a visible minority had no significant effect on workers’
attitudes toward the four perceptions of fairness of the promotion process or on their
perception that they had been denied a promotion or a developmental opportunity.

4.

Differences in Barriers Identified

Exhibit II-3 compares the barriers identified by visible minorities with those

cited by other public servants. In this case, in contrast to the pattern for aboriginal respon­

dents, ’’membership in a visible minority" (42%) and "not part of the group" (43%) were
the most frequently identified reasons for denial of promotion. Over two-fifths of visible

minority public servants gave these responses.
Visible minority respondents were also a little more inclined than public

servants as a whole to say that managers did not think they were ready (31%) or that they

would not support their application (32%). Otherwise, as observed for the two other target
groups, the barriers identified followed a profile similar to that of other public servants.

Abt Associates of Canada

18

Barriers Identified By Visible Minorities

EXHIBIT II-3

Visible minority
ALL

Yes

No

Unweighted N|Weighted N|Weighted X Unweighted N Weighted N Weighted x|Unweighted N|Weighted n|Weighted X

TOTAL DENIED
Manager did not think I was ready
Manager did not want to replace me
Manager did not give me information
Manager would not support my application
Did not get along with supervisor
Did not want overtime, set hours, shifts
I was too young
I was too old
I was or was planning to become pregnant
I am a woman
I am a man
I am a member of a visible minority
I am a person with disabilities
I, am an aboriginal person
Would not normally hire a woman
Would not normally hire a man
I had young children, dependents at home
I worked part-time or shared a Job
I was divorced, separated
I was a single parent
I objected to sexual harassment
Took maternity/patemity/parental leave
Took language/education leave
Took long term disability/sick leave
I am an anglophone
I am a Francophone
I could not relocate
I was not "part of the group"
I was unable to travel
X did not meet a language requirement
Downsizing, budget cut
Competition, promotion process not fair
Nothing available
I lack educational requirement
Other

3,878
787
941
526
809
426
68
221
197
92
404
287
48
60
8
131
28
130
60
28
29
44
91
20
52
779
203
233
1,109
53
47
50
111
294
52
985

66,083
13,217
16,424
9,824
14,898
7,573
1,229
3,603
3,309
1,298
5,072
5,543
1,093
1,168
150
1,464
543
2,053
912
485
472
641
1,423
314
1,087
12,706
3,636
3,596
19,251
874
652
777
2,123
5,191
994
16,872

100.00
20.00
24.85
14.87
22.54
11.46
1.86
5.45
5.01
1.96
7.68
8.39
1.65
1.77
0.23
2.22
0.82
3.11
1.38
0.73
0.71
0.97
2.15
0.48
1.65
19.23
5.50
5.44
29.13
1.32
0.99
1.18
3.21
7.85
1.50
25.53

226
66
48
48
69
33
0
11
16
1
29
11
96
6
3
6
4
5
2
1
3
5
1
1
■3
33
4
10
97
0
4
0
8
14
2
56

191
254
15
279
151
1,630
99
50
33
77
70
49
25
58
94
6
4
75
519
91
168
1,651

100.00
30.59
21.43
22.17
31.86
15.00
•
4.92
6.56
0.39
7.21
3.90
42.06
2.56
1.29
0.84
1.99
1.81
1.26
0.64
1.50
2.44
0.16
0.09
1.94
13.38
2.35
4.34
42.60

62

1.60

150
192
44
1,061

3.87
4.97
1.15
27.37

3,877
1,186
831
859
1,235
582

4,104
853
989
574
878
459
68
232
213
93
433
298
144
66
11
137
32
135
62
29
32
49
92
21
55
812
207
243
1,206
53
51
50
119
308
54
1,041

69,959
14,403
17,255
10,684
16,133
8,154
1,229
3,793
3,563
1,313
5,351
5,694
2,724
1,268
200
1,497
620
2,123
960
510
530
735
1,429
318
1,162
13,225
3,727
3,764
20,902
874
714
777
2,273

100.00
20.59
24.66
15.27
23.06
11.66
1.76
5.42
5.09
1.88
7.65
8.14
3.89
1.81
0.29
2.14
0.89
3.04
1.37
0.73
0.76
1.05
2.04
0.45
1.66
18.90
5.33
5.38
29.88
1.25
1.02
1.11
3.25

1,038
17,933

1.48
25.63