Demographic study of Visible Minorities in the FPS

Prepared by

Douglas Booker, Natalie Dole, Stan Lee, Kathy Malizia, Daniel O'Connor, and Rhonda Nause

For

The federal task force on visible minority participation in the public service

Final report

February, 2000

[ Acrobat Version ] [ Version française ]

Demographics, Labour Market Analysis and Modeling Unit
Research Directorate
Policy Research and Communications Branch
Public Service Commission

Disclaimer: this paper represents only the views of its contributors, and not those of the Public Service Commission.

Executive summary

This report was prepared in support of the PSC's overall objective of achieving a representative workforce and more specifically as part of a program of research supporting the "taskforce on the participation of visible minorities in the federal public service" . The taskforce was established in 1999 with the objective of providing "an action plan, aimed at improving the participation of visible minorities in the federal public service". In support of the work of the taskforce the PSC undertook a demographic analysis which provides context to the overall situation of visible minorities in the federal public service.

The report examines three key areas with respect to demographics: current levels of representation; current rates of hiring and promotion; and separation rates. The report concludes with a section on forecasting future representation of visible minorities in the federal public service. The study provides a baseline for defining some of the key issues as well as the magnitude of the challenge of achieving a representative workforce.

Key findings include:

While visible minorities are under-represented relative to labour market availability, representation has increased consistently through the 1990's. From 4.1 % in 1990-1 to 5.9 % in 1998-9.

• The increase in representation observed in the mid 90s results from a lower attrition rate for visible minorities. Essentially they were less likely to leave during the program review period.

A major barrier to increased representation public service wide, is that some of the largest departments have among the lowest representation.

• The third (National Defence), fourth (Correctional Services) and sixth (Fisheries and Oceans) largest departments had representation of 2.8%, 2.8% and 2.6% respectively.
• In some instances (e.g, Fisheries and Oceans and Indian and Northern affairs, and Heritage Canada), representation of visible minorities has actually declined in the 1990's
• Of the 20 largest departments, only two, Revenue Canada and Health Canada have achieved or surpassed labour market availability.

Profound occupational concentrations based on designated group status can be found in the public service.

• Visible minorities are highly concentrated in the scientific and professional group
• Visible minorities are more likely to be found in positions considered to be knowledge jobs 63.5% are in these positions compared to 52.6% of the non-visible minority population.

These concentrations in the public service parallel those found in the general labour market.

The low representation of visible minorities in the executive category 3.0 % and seemingly slow rates of progress over the last 8 years are areas of concern.

• Despite representation in the feeder groups that was relatively stable at 6.0% between 1991 and 1997, executive representation continues to be low.
• To some degree the under representation can be explained by low representation of visible minorities (3.7%) in the key feeder groups which are more likely to provide a career path to the senior levels.

Low levels of external recruitment of visible minorities has been the most significant barrier to improved representation. The projections demonstrate that if current rates of inflow continue that the public service will still not achieve a representative public service by the year 2023. Achieving a more representative public service (using the 1996 census as the benchmark), by 2005 would require that 1 in every 5 new recruits to the public service be a visible minority. The magnitude of this task speaks to the need for increased corporate efforts and policy initiatives to build a more representative public service.

Table of contents

• I. Introduction
• II. Demographic trends: current research
• III. Characteristics of the visible minority population in the federal public service
  • 3.1: Population and tenure
  • 3.2: Departmental and agency populations

  • 3.3: Distribution by occupational category

  • 3.4: Distribution by occupational group
  • 3.5: Executive feeder groups and the executive category
  • 3.6: Distribution by level
  • 3.7: Geographic distribution
  • 3.8: Age profile and length of service
• IV. Visible minority community mobility characteristics
  • 4.1: Recruitment
  • 4.2: Promotional activity
• V. Forecasts of attrition and future representation
• VI. Key issues and conclusions
• List of appendices
  • Appendix 1: Technical notes on the population
  • Appendix 2: Technical notes on labour market availability
• Bibliography
• Notes

I. Introduction

As part of the overall effort to address issues related to a representative public service, the Treasury Board president announced the creation of a task force on the participation of visible minorities in March 1999. The mandate of the task force is to provide "an action plan, aimed at improving the participation of visible minorities (1) in the federal public service". In order to assist the task force in the development of their action plan, forecasting and demographic research services have been provided by the Public Service Commission.

This report examines some of the key demographic vulnerabilities and challenges which currently face the public service regarding the participation of visible minorities. It seeks to provides information necessary to draw a more complete picture of visible minorities in the federal public service. The report examines key demographic issues including a profile of the current population, recruitment issues, mobility factors, departure rates and includes forecasts of expected visible minority representation in the federal public service. The demographic profile and forecasting results presented here are key elements in the identification of demographic vulnerabilities and in the development of hr planning strategies needed to facilitate progress towards equitable representation for this community.

II. Demographic trends and current research

Over the last decade, a number of research studies, both internal and external to the federal government (Friedman: Pearson Shoyama institute, 1998; Multicom, 1996; Samuel, 1996; O'Connor, Lee and Booker: 1999; IMRD, 1999), have been conducted on issues related to visible minorities in the federal public service. One of the common conclusions drawn from these studies, is the existence of major challenges which need to be met in order to achieve a representative work force. It has been suggested that the data indicate an 'employment equity crisis in the federal public sector.' (Friedman: Pearson Shoyama institute, 1998)

Recent forecasts on representation (O'Connor, lee and Booker: 1999) indicate that at current levels of recruitment the public service will never achieve labour market availability (Section V). In addition, several research studies (Multicom, 1996; Samuel, 1996) and the recent decision of the Canadian Human Rights Tribunal have concluded that visible minorities have faced systemic discrimination within the federal public service.

Studies have also indicated that barriers exist for designated groups within the public service. A joint PSC-TBS evaluation of employment equity special measures programs concluded that there "has not been sufficient effort made to identify and remove barriers to achieving employment equity in the federal public service" (psc:1992:i).

John Samuel & Associates Inc.'s 1996 study on the status of visible minorities in the public service used surveys, interviews and focus groups to access more qualitative information about this particular designated group. They concluded that visible minorities are not doing well at various levels of the public service, in comparison to both the federally regulated sector and the private sector. They suggest several barriers facing visible minority employees, including, but not limited to the following:

• A lack of flexibility in the staffing process;
• A perception that candidates for appointments and promotions are 'pre-selected';
• Difficulty in getting information on jobs;
• A level of discomfort with the workplace Environment and a feeling that networking, kinship and social ties play a major role in hiring and promotion;
• An overall negative perception of the PS by the visible minority community, exacerbated by cultural biases and low salary levels for some professions in comparison with the private sector;
• The lack of a simple and structured approach to career development and training opportunities;
• A failure to focus efforts to hire visible minorities into senior management;
• A need for diversity training opportunities; and
• A perception that PS managers are not fully committed to increasing visible minority representation.

Special attention was given to the flexibility and accountability of hiring policies and procedures, "restrictions on external recruitment are insufficiently flexible to provide increased employment opportunities for visible minorities" (Samuel, 1997) as well as the lack of accreditation given to foreign education and languages, feelings of exclusion from the PS 'corporate culture', as well as biased interview practices. It was suggested that the only way for employment equity goals to be reached would be for commitment for a new diversity strategy at the highest level of authority. Barriers must be removed to ensure that real merit is recognized and rewarded.

Samuel suggests that other factors contributing to low visible minority employment is the nature of public service work, competition from the private sector, the fact that some visible minorities seek employment in very specialized fields and that some departments are located in regions with small visible minority populations. (Samuel, 1997).

One of the major criticisms aimed at the Samuel study was that it was based primarily on anecdotal evidence or small sample sizes. While some of the issues identified by Samuels have been confirmed in follow-up work (IMRD 1999) there still remains a critical gap in the research in that there has been no systematic evaluation of the "demographics" of the issue. The objective in this paper is to fill this gap with a comprehensive profile of key demographic issues.

III. Characteristics of the visible minority workforce in the federal public service

3.1 Population and Tenure

According to the 1996 census report, representation of visible minorities in the Canadian labour market available to work for the federal public service, was 8.7% (2) as of March 31, 1999 visible minorities represented only 5.9 % (3) (see table X) of the public service population (10,586 employees excluding separate employers). The representation of visible minorities in the indeterminate population of 6.0 % (n=9,171) was significantly higher than their representation in the term population (5.5 % : n=1,415). (4)

While visible minorities remain under-represented, representation has increased consistently through the 1990's. From 4.1 % in 1990-1 to 5.9 % in 1998-9. It is important to note, however, that this increase occurred through a lower attrition rate of visible minorities in the 3 first years of program review compounded with a recent increase in their recruitment share and departmental efforts made to improve self-identification data (1997-8 and 1998-9). In fact, although there has been a steady increase in the representation of visible minorities through the 90s, in raw numbers the visible minority workforce presents a varied history. As table X reveals the public service has experienced both increases and decreases in the number of visible minorities through the 90s. The early nineties are marked by a steady increase in the number of visible minorities going from 9,815 at end of fiscal 1990-1 to all time highs of 10, 630 and 10613 in 1992-3 and 1993-4 respectively. The number of visible minorities decreased gradually thereafter to an a low of 9525 at the end of fiscal 1996-7 only to surge drastically in 1997-8 (10,809) and 1998-9 (10586).

The increase in visible minority representation observed in the mid 90s results from a lower attrition rate of visible minorities. In fact, from 1994-5 to 1996-7, the visible minority population decreased by 9.3 % compared to an attrition rate of 15.3 % for non-visible minorities. From 1996-7, although the non-visible minority population decreased by a further 8.3 %, the visible minority population increased by 11.1 %. The bulk of the decrease in the visible minority population between 1990-1 and 1996-7 was in the term population as the indeterminate population actually increased slightly from 8,004 to 8,154.

Within the visible minority population, the representation of women (50.5 %) is higher than that of men (49.5 %). This is slightly lower than the representation of women in the non-visible minority group where they represent 51.7 % of the workforce. The gain in the representation of visible minority women over the 90s is considerably higher (16.3 % increase) than the gain in representation for visible minorities as a whole (7.8 % increase). On the down side, the progress in representation made by visible minority women (46.8 % in 1991 to 50.5 % in 1999) is lower than that of non-visible minority women (46.8 % in 1991 to 50.7 % in 1998).

3.2 Departmental Populations

The ten largest federal departments (table 3.1) employ 67.6 % of all public service employees. The largest department is Revenue Canada which comprises 20.9 % of all public service jobs, followed by Human Resources Development Canada (11.6%) and the DND (8.7 %). the largest populations of visible minorities at the departmental level (table 3.1: distribution) are found in Revenue Canada (35.3 %), HRDC (9.1 %) and Public Works (6.7 %).

Table 3.1 indicates there are great differences in the representation of visible minorities at the departmental level. Currently the best representation among the 20 largest federal departments is found at Citizenship and Immigration (not on the table) and Health (9.9 %), followed by Revenue Canada (5) (9.1 %) and Statistics Canada (7.6 %). Interestingly, Health Canada was one of the departments with the highest representation in the federal public service even before the impacts of the 1997 tribunal decision could fully be considered a factor.

The critical measure for departments is actually their performance against labour market availability (LMA) (tables 3.1 and 3.2). The best performing departments include: Health (19.3 % above availability), Revenue Canada (13.8 % above availability), Citizenship and Immigration (98% of availability), Corrections (92.3 % of availability), Statistics Canada (89.4%), and Public Works (89.3 %).

The lowest ratios of population to availability are found in Fisheries and Oceans (57.8 %), Environment (63.3 %), HRDC (64.4 %), RCMP (66.2 %), and Transport and Industry (67.1 %). the list of departments with the largest gaps includes the second and sixth largest departments, (now the first and fifth largest with the departure of Revenue Canada) which have a significant, albeit negative, impact on overall public service performance. Among the 20 largest departments the most significant progress has been made by Environment, Transport, and Foreign Affairs where representation increased by 93.8 %, 82.1 %, and 64.5 % respectively between 1990-1 and 1998-9 (table 3.2). It is important to note that of the 20 largest departments, only Revenue Canada and Health have achieved or surpassed labour market availability and that in some instances (e.g, Fisheries and Ocean and Indian and northern affairs, and Heritage Canada), representation of visible minorities has dropped (6).

One can conclude that efforts need to be more focused at the departmental level particularly among departments who appear to be either stalled or regressing in terms of progress and among those whose performance relative to labour market availability is very weak. In order to improve departmental performance, it may be useful to examine best practices of departments which were successful in increasing representation of visible minorities and determine if these practices can be transferred to other departments.

3.3 Distribution by Occupational Category

Profound occupational concentrations based on designated group status and gender remain in the public service. Visible minorities are highly concentrated in the scientific and professional group with 23.6 % of all visible minorities in the public service are found in this group compared to 11.9% of all non visible minority public servants. There is a strong gender dimension to this representation in the scientific and professional category as only 13.3 % of visible minority women are found in the scientific and professional jobs versus 34.1 % of visible minority men.

Overall the visible minority population within the public service occupy a higher proportion of what would be considered the knowledge jobs in the public service (63.5%) than do non-visible minorities (52.6%) (7). The primary area of concern within occupational categories is the executive group where the distribution of visible minorities was about half of what it was in the rest of the population.

Representation at the category level remains a major concern. The greatest difference between availability in the labour market and public service population could be found in the technical and executive categories where representation was less than 50% of availability. There is also a considerable gap in the administration and foreign service and administrative support categories where representation was about 30% below availability the only category to surpass labour market availability is the scientific and professional category which is currently 0.5 % above labour market availability. This indicates that further recruitment efforts are required in all but one category to ensure a representative public service.

3.4 Representation by Occupational Group

The table 3.4 depicts the top 10 occupational groups with the largest visible minority population. Among these occupational groups, representation in the PMS, the AUS, the ENS, and the ESS exceed availability in the external labour market. Again, within these groups, the lowest representation relative to LMA is found in the EGS (34.4 % of availability) and the ass (42.9 % of availability).

3.5 Executive feeder groups and the executive community

The low representation of visible minorities in the executive category 3.0 % and seemingly slow rates of progress over the last 8 years (table 3.5) warrants further inquiry. Visible minority representation in the feeder groups (the EX equivalent, EX minus 1 and EX minus 2 levels) remained relatively stable at about 6.0% between 1991 and 1997 followed by an increase of 0.4 % (6.4 %) in 1998 and 1999. This indicates that there have historically been sufficient numbers in the executive feeder groups to support growth at the executive level. The key problem here is that the executive community is drawn not from the entire range of feeder groups but rather from a select number of feeder groups. An earlier study by the PSC (Malizia and Booker, 1998) revealed that only 10 occupational group levels were the source for about 70.0% of all internal movements into the executive group (PM-06, FS-02, AS-07, ES-07, PM-05, CO-03, AS-08, ES-06, IS-06, and FI-04).

To some degree the under-representation is evident for visible minorities at the executive level and is also reflected in the representation rates of visible minorities among these top ten EX feeder groups. Overall, the study revealed that in 1998 the proportion of members of visible minorities in the top ten feeders was 3.7%, greater than that of the EX group (2.6%), but smaller than the total indeterminate public service (5.1%). Averaged over seven years, representation of visible minority members among new entrants to the executive community from the key feeder groups was 2.5%, compared with 3.4% from all EX feeder groups and 6.8% from external recruitment (i.e., recruitment outside the PS). As with aboriginal peoples (Malizia & Booker 1999), external recruitment to the executive level may be an avenue worth examining and utilizing more fully to increase and enhance representation of visible minorities in the EX community.

Another dimension to this issue is that the concentration of visible minorities in the scientific and professional groups may restrict opportunities for movement into the executive ranks. Here the more common career progression appears to be in the higher echelons of the specific scientific and professional groups. As a result we see a high representation of visible minorities in the executive equivalent classifications (table 3.6).

3.6 Distribution by level

In table 3.7, the representation of visible minorities at each level of the executive category is presented for both 1991-2* and 1997-8. Highlighted levels represent levels where the representation of visible minorities is greater than their representation in the category as a whole (indicating potential compression issues).

Representation of visible minorities in the executive category has increased by 0.6 % from 1991-2 to 1998-9. This improvement is mainly due to the differential attrition rates of non-visible minorities (NVM) and visible minorities during program review. As the table illustrates, between 1991-2 and 1997-9, the executive category workforce was reduced by 26.2 % (-1105), 99.3 % of which were NVM yielding an attrition rate of 26.6 % for NVM compared to 8 % for visible minorities.

Table 3.7 shows the compression of visible minorities at lower levels of the executive category is less profound today than it was in 1991-2. In 1991-2, 91 % of visible minority executives were found in the EX 1 and EX 2 levels compared to 81.1 % for NVM (difference of 9.9 %). In 1998-9, the percentage of visible minority executives in the EX 1 and 2 levels was down slightly to 84.8 % compared to 76.8 % for NVM (difference of 8 %).

The following table (3.8) depicts the representation of visible minorities in the four largest occupational groups by level and overall (it is important to note that these groups alone comprise 53.6 % of the term and indeterminate public service workforce). Of these groups, only the pm group has achieved or surpassed LMA representation.

For some occupational groups, the compression of visible minorities in lower levels remains a concern. In the table, visible minority representation at each level is shown. Highlighted cells represent levels in which the representation of visible minorities exceeds that of their representation in the occupational group as a whole (indicating potential compression issues).

It is difficult to make a case for systematic compression of visible minorities at low levels within occupational groups. There are clear differences across occupational groups with regard to the issue of compression. Looking at the four largest groups which covered over 50% of the public service population there is evidence of compression only in the CS community where representation from the CS-02 level and up decreased consistently at each level. There was also some evidence of compression in the PM community where representation of visible minorities was clearly highest at the PM-01 and PM-02 levels and lower at higher levels. In the as and CR groups there is no evidence of compression. In fact, representation at higher levels in these groups (AS-05 and above) is generally higher than at the lower levels. However, comparisons of employment equity groups representation across levels should also include a comparative analysis of years of service to avoid penalizing departments for recent efforts in recruiting designated group members.

Table 3.9 depicts the representation of visible minorities by level in the five occupational groups with the highest representation of visible minorities (8). Except for AUS and SGS where they are compressed in the lower levels, there is a tendency for visible minorities to be concentrated in the intermediate (CHS, ENS, and SES). The compression of visible minorities in the first and second levels of the au (9) and SG group is particularly problematic given that their representation generally decreases with increases in level.

3.7 Geographic Distribution

Table 3.10 presents the population of visible minorities by region along with regional representation and relevant LMA. It is evident that visible minorities and non-visible minorities are distributed differently across regions, there is a greater concentration of visible minorities in particular regions. In fact, of the 14 provinces, territories, and regions (including external) that comprise the public service, 78.8 % of visible minorities are found in three regions, the National Capital Region (32.6 %), Ontario (27.0 %), and British Columbia (19.1 %), while only 59.5 % of the non-visible minority workforce are found in these locations. Large regional shares of visible minorities do not necessarily translate into adequate representation. Of these three regions, only Ontario has surpassed or achieved labour market availability.

Overall, six of the 13 regions have achieved or surpassed their LMA target (Ontario, Prince Edward Island, Nova Scotia, Saskatchewan, Alberta, and the Northwest Territories). Excluding locations outside Canada, the regions presenting the greatest difference between actual and expected representation are Newfoundland (57.1 % of availability), Yukon (60.9 % of availability), and Québec (68.8 % of availability). Of the remaining regions, the three regions closest to LMA are British Columbia (96.6 % of availability), national capital region (81.2 % of availability), and new Brunswick (80.0 % of availability).

Between 1990-1 and 1998-9, all regions, except Newfoundland, Prince Edward Island, and new Brunswick, have shown some progress in the representation of visible minorities with Québec presenting the greatest increase, followed by British Columbia and Ontario.

3.8 Age Profile and Length of Service

Internal public service demographics for visible minorities and NVM illustrate different profiles with respect to age distribution and distribution by pensionable years of service (see chart 3.2 and 3.3). Although, on average, visible minorities are slightly older (43.3) than NVM (42.7), they are more evenly distributed across age groups.

These differences in the age distribution of visible minorities and NVM are also reflected in their respective distribution by pensionable years of service (chart 3.3). As can be seen in figure 3,3 visible minorities have a comparatively higher concentration of the workforce in the 0 to 12 years of pensionable service whereas NVM tend to be overly represented in the 18 to 35 years of pensionable service.

These are important differences as the expected departures due to retirements over the next 10 years will be qualitatively and quantitatively different for visible minorities and NVM. The high concentration of the NVM workforce in the 40 to 50 year old interval compounded with their comparatively higher representation in the 18 to 35 years of pensionable service is projected to accelerate the rate of retirement departures over the next 5 to 10 years. For visible minorities, the mix of age distribution and distribution of pensionable years of service is expected to remain stable at a lower rate of retirement departures over the same period. Additionally, because visible minorities tend to have a lower retirement probability profile than NVM, their accumulated percentage of retirements over the same period should be lower than that of NVM.

IV. External Recruitment

4.1 Recruitment

The two key inflows into the federal public service population are the recruitment of term and indeterminate employees (10). The term population has become an increasingly important source of inflow in recent years. A growing proportion of the indeterminate population has been drawn from the term population. Any strategy designed to improve representation must focus on recruitment at both levels. The low levels of external recruitment of visible minorities in these areas represents the most significant barrier to improved representation. Between 1990-99 the visible minority share of recruitment was 7.1 %, significantly below labour market availability (see appendix 1: technical note 8).

The overall trend in recruitment over the last nine years is an area of concern. Aside from fiscal 1998-1999 where the recruitment share of visible minorities increased from the previous year, the 90s showed a steady decline in the proportion of visible minority external term and indeterminate recruits.

Recruitment of visible minorities dropped dramatically during the program review years of 1993-4 to 1997-8. Despite increased efforts to promote employment equity, public service wide recruitment levels have only exceeded 4 percent in 1999 over the last 6 years. Performance was considerably better in the 1990-1 to 1992-3 period when recruitment levels averaged 6 percent of total intake per year.

In 1998-9, only in nova Scotia did the recruitment share of visible minorities exceed regional labour market availability. The highest levels of visible minority recruitment at the regional level are found in British Columbia (8.6%) vs an LMA of 11.8 % , ontario(6.6 %) vs 9.6 % and the NCR (6.1 %) vs 6.9 %,

Despite this relatively poor performance public service wide, at the departmental level there are some success stories worthy of mention. Statistics Canada recruited new visible minority entrants into the public service at 19.6% over the nine year period, with recruitment exceeding 10% each year. In addition, the department of veterans affairs achieved a 9 year average of 3.9 % despite being located in a labour market (PEI) where representation was only .5%.

Table 4.3 above, depicts the trend in visible minority recruitment shares for the 20 largest departments over the period of 1990-1 to 1998-9. As can be seen, despite the overall increase observed in the recruitment share of visible minorities in 1998-9, decreases in the share of recruitment for visible minorities are observed for all but 4 of the largest departments. Among these 20 largest departments the most significant decreases in recruitment share for visible minorities have been experienced by, Corrections (81.8% over 8 years), Heritage (80% over 4 years, and Fisheries and Oceans (72.3% over 8 years). In 1997-8, of the 20 largest departments, only Statistic Canada and Foreign Affairs had recruitment shares which exceeded their respective LMA for visible minorities.

4.2 Promotional Activity

At the category level, the promotion rates of visible minorities exceeded those of non-visible minorities between 1990-1 and 1997-8. Visible minorities were promoted at rates exceeding non-visible minorities in every occupational category (table 4.4).

The promotion rates of visible minority women exceeded those of visible minority men in all occupational categories between 1991-9. These findings indicate that over the past98 years visible minority women have been promoted at levels above the representation in the public service and at higher rates visible minority men and non-visible minorities .

The findings indicate with respect to promotions that there are no service wide issues (at least at the aggregate level). Relative to there representation in the public service visible minorities receive promotions at levels equal to or possibly exceeding non-visible minorities, though issues may exist at the departmental or occupational group level. The promotion issue does not appear to constitute a public service wide issue except perhaps at the executive level where inflow into the executive community has been consistently lower than internal representation in the feeder groups over the last decade

It should be noted however, that the analyses on promotion rates presented in this paper are conducted at an aggregate level (i.e., category level), it is possible that a more refined analysis would reveal lower promotion rate for visible minorities at other levels (e.g.., occupational groups). Because of this, the data presented in this paper, do not discount the possibility of specific problem areas at the departmental or occupational group levels.

V. Forecasts of attrition and future representation

Forecasts are achieved by assessing historical patterns of mobility (i.e. recruitment, separations, and retirements) for each employment equity group. These mobility patterns are then used to predict the future behaviors of employment equity groups. This strategy is particularly well suited for representation issues as representation is greatly affected by mobility patterns (i.e. inflows and outflows).

The forecasting model used is primarily a 'pull' model, that is, the model is designed to forecast departures from the public service. The projected departures create vacancies 'a vacuum' that acts to draw in new recruits. The new recruits are then divided between different employment equity groups and the non-equity group, based on their respective historical recruitment share (1991-1998).

In order to simplify forecasting, it is assumed that the current public service population will remain steady over the course of the simulation which runs from 1999 to 2023. The steady state is achieved by adjusting recruitment to match the expected number of departures. That is, the model assumes all forecasted vacancies will be replaced.

Recent forecasts have demonstrated that, although some progress will be made over the next 10 years, visible minorities will continue to be under-represented without substantial efforts to improve recruitment levels (O'Connor, Lee and Booker 1999).

Historically (1991-2 to 1994-5), data used to establish exit patterns for visible minorities show overall separations rates and retirement probabilities which are lower than those of other employment equity groups or the public service as a whole.

As can be seen in table 4.6., from 1992 to 1995, separation and retirement rates were 2.25 and 6.60 percent annually. Comparatively, these figures are substantially lower than what is observed, overall, in the public service where separation and retirement rates are 3.01 and 15.10 percent respectively.

From a forecasting perspective, such low exit patterns for visible minorities point to recruitment as the main obstacle for achieving 1996 census labour market availability (LMA) representation. In fact, although using their historical recruitment share of 7.1 %, visible minorities are not expected to achieve LMA representation within the next 20 years, their representation is expected to increase unabated over the same period (see chart 5.1). Given a historical recruitment share (7.1 %) which is below their LMA (8.7 %), this increase in representation for visible minorities is, in great part, attributable to their low rates of attrition.

As far as the representation of visible minorities goes, it is unlikely that much more can be achieved through retention. Indeed, as shown in chart 5.1 and in the initial forecast using historical rates retention for visible minorities is not an issue. There is, however, much more room for policy changes at the level of recruitment.

Recruitment scenarios : assessing hypotheses chart 5.2 depicts forecasted visible minority representation using three different goal oriented recruitment scenarios. In the first scenario (line with diamond markers), recruitment share is set at LMA (8.7 %). In the second and third scenarios, the goal of achieving visible minority LMA representation by 2010 and 2005 is implemented by increasing recruitment share for visible minorities accordingly (see chart 5.2).

For visible minorities, a recruitment share set at LMA is only expected to result in LMA representation by the year 2019. In terms of share of new recruits, increasing the recruitment share to 8.7 percent from the historical share of 7.1 percent, implies that over the 1999-2019 time period, 1 in of every 15.6 recruits will be a visible minority.

In order to set more aggressive goals with respect to the time frame (i.e. time to achieve LMA representation), recruitment must be increased accordingly. As can be seen in table 5.2, LMA representation can only be achieved by the year 2005 if recruitment is increased to 20.1 % attributing 1 out of every 4.9 new recruits to visible minorities.

The forecasts illustrate the key problem areas for related to representation of visible minorities in the federal public service. The gap between workforce representation and labour market availability is large. According to our analysis, retention is not a serious issue for this group, and, according to our projections, current recruitment efforts will be insufficient to close the gap within the next 20 years. Even pegging recruitment levels to the rate of availability in the external labour market will not overcome the differences between the internal and external markets in a reasonable length of time. Realizing the goal of a representative workforce will require serious revisions to current practices as recruitment efforts will have to be two to three times current levels to reach our corporate objectives.

VI. Issues and Conclusions

The critical issue identified by this demographic analysis is that visible minorities are under-represented in the public service.

Relative to their representation in the public service, visible minorities receive promotions at levels equal to or possibly exceeding non-visible minorities. It should be noted however, that the analyses on promotion rates presented in this paper are conducted at an aggregate level (i.e., category level), it is possible that a more refined analysis would reveal lower promotion rate for visible minorities in some areas (e.g.., occupational groups). Visible minorities also tend to leave the public service at lower rates than non-visible minorities and work longer after becoming eligible to retire than the non-visible minority population.

Low levels of external recruitment of visible minorities has been the most significant barrier to improved representation. The projections demonstrate that if current rates of inflow continue that the public service will still not achieve a representative public service by the year 2023.

While departments have shown different levels of success with respect to their performance relative to labour market availability only two departments (Health and Revenue) had achieved this "target " by March 31, 1999. This in itself speaks to the need for increased corporate efforts and policy initiatives to build a more representative public service. At the departmental level, however, efforts need to be more focused particularly among departments who appear to be either stalled or regressing in terms of progress and among those whose performance relative to labour market availability is very weak. Efforts clearly need to be focused on recruitment of new talent if progress is to be made. In this area many departments appear to be labouring as recruitment levels in many departments seem to have fallen since the early 1990's. The positive examples of Health, Revenue, Statistics Canada and Veterans Affairs need to be emulated in order to achieve the needed progress towards a more representative public service.

Appendices

Appendix 1: technical notes on population and appointment information

1. Population data is collected as of March 31st in each fiscal year.

2. Appointment and separation data is collected throughout each fiscal year beginning April 1 and ending March 31st.

3. The population, appointment and separations data is collected on all departments and agencies within the Public Service Commission universe as defined under the public service employment act (PSEA). This includes: civilian personnel only in the department of national defence and the RCMP. The PSC universe does not include: crown corporations (Canada post, CMHC) or any of the newly created agencies (NavCan, CFA)

4. There are slight differences between the pac and t's universes which account for some differences in overall population.

5. Differences in the treatment of information between the PSC and the TBS will produce differences in population counts, for example the PSC only uses substantive positions when compiling information, therefore someone on assignment outside there department or at a higher level is identified only in there substantive position and the department in which that position resides. TBS in compiling data counts acting or interdepartmental assignments.

6. Separate employers (i.e. Indian oil and gas, superintendent of financial institutions) are not included in the analysis as these organizations are not required to collect self-identification data.

7. The term population as defined under the PSEA includes all term employees over 3 months in duration. Prior to 1993 the term population included only term employees of over 6 months in duration.

8. Figures used in this report are higher than PSC annual report numbers, because of delays in self-identification reporting. There is a significant time lag between the appointment effective date of new employees and the entry of self-identification data. The majority of appointments are captured within 9 months of the end of the fiscal year. PSC data is normally reported based on the self-identification counts at the end of the fiscal year. For the purpose of this analysis all persons who have self-identified and are in the TBS-EEDB irrespective of when the data is captured are back dated to their appointment effective date.

Appendix 2: some notes on labour market availability, labour force data

Labour market availability is the legal basis for the definition of under-representation in the federal public service. Labour market availability in Canada for visible minorities is 8.7%. This figure is derived from population information collected in the 1996 census. This figure differs from region to region based on the demographics of regional labour markets.

The actual labour market availability figure is drawn from labour force participation rate of visible minorities of 10.3%. The difference between the overall participation rate and the la figure is based on two key factors:

1. The matching of positions in the labour force to public service jobs through the national occupational classification or noc codes and
2. through the exclusion of non citizens from inclusion in the labour market availability numbers.

Bibliography

• Cronshaw, Steven f. "External review of selected instruments of the PSC's personnel psychology centre, organization and management solutions (for the information management and review directorate), Public Service Commission, may 1999
• Information Management and Review Directorate (IMRD), "Employment systems review: PSC's general recruitment and priority administration systems", Public Service Commission, may 1999.
• Malizia, Kathy and Douglas Booker "demographic profile of key executive feeder groups, 1991-1998" PSC, research directorate, 1999.
• Multicom (1996) a consultation with visible minority employees at Canadian Heritage. Ottawa: dept of Canadian Heritage.
• Nehme, Micheline "demographic profile of the federal public service for the fiscal years 1994 - 1997" Research directorate, Public Service Commission, 1998.
• O'Connor, Daniel, Stan Lee and Douglas Booker "Toward a convergence of internal and external labour markets: Forecasting employment equity representation in the federal public service 1998-2023" (forthcoming) research directorate 1999.
• Public Service Commission of Canada (1996) annual report 1995-96, minister of Public Works and Government Services Canada.
• John Samuel and associates (1996) visible minorities and the public service of Canada. Ottawa: a report submitted to the Canadian human rights commission.
• Treasury Board secretariat, employment equity council (Canada) and visible minority consultation group. Distortions in the mirror: reflections of visible minorities in the public service of Canada. Report of the visible minority consultation group to the secretary of the Treasury Board and the employment equity council of deputy ministers. Ottawa: visible minority consultation group.

Notes

1. A visible minority is defined as anyone (other than an Aboriginal person) who is non-white in colour/race, regardless of place of birth. This includes persons from the following visible minority groups or origins: Black, Chinese, Filipino, Japanese, Korean, South Asian/East Indian (including Indian from India; Bangladeshi; Pakistani; East Indian from Guyana, Trinidad; East Africa; etc.) Southeast Asian (including Burmese; Cambodian; Laotian; Thai; Vietnamese; etc.), Non-White West Asian, North African or Arab (including Egyptian; Libyan; Lebanese; Iranian; etc.) Non-White Latin American (including indigenous persons from Central and South America; etc.), Person of Mixed Origin (with one parent in one of the visible minority groups listed above); other visible minority group.

2. Labour Market Availability data is calculated from the 1996 Census. Population and Appointments data from TBS and PSC will show slight differences (See Technical Notes Appendix 1).

3. The employment equity data used in this report based on voluntary self-identification data collected by departments and maintained through the Treasury Board's Employment Equity Data Bank (EEDB).

4. This lower levels of representation in the term population are related to the administration and processing of the self-identification questionnaire and data. There is often a six to nine month lag in the entry of self-identification information into the TBS- EEDB. For more detail see Appendix 1 Technical Note 8.

5. When Revenue Canada is removed from the public Service the overall indeterminate population drops to 125,623. Overall representation of visible minorities in the indeterminate population drops from 5.1% back to 4.5%.

6. . The declines in visible minority population in certain departments may result from the reorganization of departments, this issue requires further examination.

7. Knowledge work defined as working in professional categories, either Executive, Scientific and Professional or Administration and Foreign Service.

8. The VM group presented the highest visible minority representation with 18.8 %. However, there are only 32 VMs overall and with such small numbers, a difference of 1 individual at a given level can increase or decrease representation dramatically. Because of this, the VM group was not included.

9. It should be noted that the majority of AU's are employees of Revenue Canada and will no longer be considered part of the Public Service following the transfer of this department to Agency status.

10. It is difficult to track the designated group status of the term population, because of the lag in the collection of equity data (See Appendix A: Technical Note 8) and because of the short duration of many of these term appointments makes it difficult to collect Employment Equity data as the may only have a single opportunity to self-identify.