Dr Martin Dunn

Recruitment and Promotion in the Australian Public Service

This submission relates to recruitment and promotion within the public sector. The Public Service Act 1999 section 10(1)(c) declares that “decisions relating to engagement and promotion are based on merit”. This declaration has its origins in the Northcote–Trevelyan Report of 1854 that introduced the basis of the modern professional public service in the United Kingdom. Despite this, the way recruitment and promotion is managed in the Public Service does not guarantee that the most appropriate applicant is selected and is vulnerable the unconscious and deliberate bias.

While there can be variations on the selection process, the standard process involves the following steps:

the position is advertised with applicants referred to a set of generic selection criteria;

applicants are screened and a shortlist is compiled based on their written applications;

applicants on the shortlist are interviewed;

a referee check is conducted for the applicants considered most suitable;

the most suitable applicant is offered a position. A merit list may be retained with the remaining suitable candidates to use for filling vacancies that arise

Each of these steps is problematic.

Generic selection criteria. Senior Executive Service staff are recruited against the Senior Executive Leadership Capability Framework (introduced in 1999) while more junior staff are recruited against the Integrated Leadership System (introduced in 2004). While generic criteria offer an opportunity for standardisation, they do not accurately reflect the requirements of the positions. The main issues are:

Domain knowledge. While main public service positions can be filled by intelligent people, who learn about the field on the job, domain knowledge is not irrelevant. As a minimum, domain knowledge helps avoid the learning curve in the first few months associated with any job, but also provide a depth of expertise that allows them to better place new problems in context. Even with senior managers, domain knowledge remains relevant. Stoller (2016) observes that better run hospitals have doctors as CEOs rather than generalist managers. Domain knowledge brings a better understanding of context and better credibility.

Numeracy and Data Skills. Numeracy and data skills are not included in the generic criteria, yet we are increasingly in an environment that employs and interprets data – most obviously on performance and costs. The APSC states that “data skills are essential for all APS employees to support evidence-based, informed decision making” (https://www.apsc.gov.au/data-literacy-skills) yet while generic criteria specify the need for communications skills there is no mention of data skills. If these skills are indeed essential (and they are) then they should be reflected in the generic selection criteria. It is worth noting that when the Senior Executive Leadership Capability Framework was being developed, Google was still operating out of a garage.

In practice, many work areas interpret the generic selection criteria in terms of what they understand as the job requirements, but the problem with this approach is there is a corresponding loss of transparency. Job applicants do not understand the real requirements of the positions and it may be that different people involved in conducting the selection process have differing interpretations.

Screening. The screening step is increasingly important in the public service as the numbers of applicants for each position has grown. It is not unusual for a position to have over 100 applicants, yet the screening step tries to reduce the number of people that need to be interviewed to a handful. The reason for growing numbers of applicants is unclear, but appears to be a feature of recent restrictions on public service growth and the use of generic criteria. The result is that plausible candidates are excluded based solely on an assessment of their written application. The problems with the screening process includes:

Assessing Applications and Transparency. It is not clear that the broad set of skills based in thebehaviours sought in the generic criteria can be assessed from a written application. At best, they can identify whether these skills are well described in the application. Inevitably, the assessors need to apply such as the nature and variety of positions. However it is addressed, much of it involves subjective judgement and the process lacks transparency.

Professional resume and application services. There is an extensive market in development of professional resumes. As a result, it can be unclear whether the applicant prepared his/her own application. Efforts to assess written skills based solely on the application may be mistaken.

At the same time candidates can be screened out for relatively minor errors, such as spelling mistakes (Martin‐Lacroux, 2017).

Bias and Familiarity. The screening process is vulnerable to the subjectivity and biases of those assessing the applications.

The Behavioural Economics Team in the Department of the Prime Minister and Cabinet recently conducted a study designed to determine if there was bias against minorities and women in the selection process – specifically determining the impact of names which revealed the applicants’ gender and ethnicity (HIscox et al, 2017). In the context of the trial, women with indigenous names were most likely to be shortlisted, while men with Anglo-Celtic names were least likely. Whether these results reflect the limitations of the trial, or a prevalence of “politically correct” bias is unclear, but the results do show that the resume screening process is vulnerable to biases and the result can be significant.

In principle, this bias can have any number of dimensions – stereotypes about particular agencies, educational institutions, roles, etc. The absence of objective ways of determining whether a particular set of soft skills is well developed encourages such stereotypes.

The most obvious form of bias is that created by familiarity. Applicants that are known to the assessor may be seen as lower risk or worth a chance. Yet this is in practice no different to the system of patronage that the Northcote–Trevelyan reforms were intended.

Acting Opportunities. The distribution of acting opportunities has a distorting impact on the selection process. Some individuals will have been given opportunities to act at a higher level because their immediate superior takes extended leave or departs unexpectedly, they have the support and patronage of senior officers and their agency does not fill vacancies promptly using a selection process. Others miss out. The individuals that have acting opportunities gain in three ways (1) acting opportunities give them actual skills in working at a higher level, (2) time in acting opportunities are one of the few indicators of capacity to work at a higher level that can be observed in a résumé; and (3) they signal the confidence that senior management has in the individual. But, the process of providing acting opportunitiesis rarely transparent – and often has more to do with the individual having fortunate circumstances rather than natural ability.

Selection Interview. The interview is used almost universally within the public service. Yet there is considerable research on selection techniques and the interview has a number of limitations. In this regard the public service is not unique in neglecting research findings on better process (Le et al, 2007).

Validity of Selection Interviewing. Research on selection techniques show that there is no ideal technique, and while the interview is one of the better techniques, it is not the best.

The consensus is that the best single predictor of future job performance general mental ability (or IQ) (Kuncel, 2010; Menkes, 2005). These can be easily administered by an online test. While intelligence does not capture every aspect that is relevant to the workplace, other assessment techniques do no better.

A combination of techniques can achieve better reliability than general mental ability alone (eg a combination of general mental ability and the interview), but the gain is relatively small.

Some academics think that the validity of interviewing may be quite low. Moore (2017) offers the conclusion that “the traditional face-to-face job interview is terrible at predicting job performance”. The reason is that the comparison between interview performance and subsequent work performance is often assessed by the same individuals, the line manager. There is a psychological barrier to admitting that one’s original assessment is wrong. This effect is visible when we look at the effectiveness of interviewing in assessing performance in training (where there are objective assessments) compared to the workplace. In these environments, interviewing does comparably worse.

The relatively poor performance of interviews has number of potential causes. This includes the propensity of interviewers to form pre-conceptions prior to the interview – either through previous experience with the candidate or in the screening process – and then (in a classic example of confirmation bias) focus on confirmatory evidence in the interview while discounting inconsistent evidence (Dipboye, 1982; Dougherty et al, 1994)). Applicants, of course, go to some length to present themselves in a positive light, and some are better at managing the interview than in actually doing the job (Barrick et al, 2009)

In terms of outcomes, as well as not obtaining the best applicants, employment interviews encourage self-promotion and serve as an environment where narcissists can do better (Paulhus et al, 2013).

Structure and Training. The academic literature suggests that interviewing is most valid when the interviews are structured and the interviewers are trained. While the typical public service interview is structured, most interviewers have not been trained in the role. There is considerable scope to introduce better practice into the conduct of public service interviews (Campion et al, 1997).

Referees. Generally, the public service does not put a lot of weight on referee reports. They are only collected for a few (most promising) applicants and the collection is often delegated to a scribe. The perception is that referees are often considered unreliable: if they like the employee they will provide a positive assessment; while if they dislike them then they will also provide a positive assessment in order to get rid of them. These views are probably unnecessarily cynical.

My suggestions for improvements are:

A project to improve selection methods should be established.

The Senior Executive Leadership Capability Framework and the Integrated Leadership System should be reviewed in light of the current work environment. As a minimum, domain specific skills and numeracy/data skills should be an obligatory part of any selection process.

Screening should be routinely be done by online testing, covering as minimum general mental ability, literacy and numeracy. A whole of government contract should allow for greater efficiency. (Note that a reform proposed by Northcote–Trevelyan to eliminate patronage was the civil service exam.)

Acting opportunities of more than a few weeks duration should be filled on merit.

A program of training for interviewers should be developed.

Greater transparency should be provided to applicants on how shortlisting would be conducted and how interview panels will assess performance.

References:

Barrick, M. R., Shaffer, J. A., & DeGrassi, S. W. (2009). What you see may not be what you get: relationships among self-presentation tactics and ratings of interview and job performance. Journal of Applied Psychology, 94(6), 1394.

Campion, M. A., Palmer, D. K., & Campion, J. E. (1997). A review of structure in the selection interview. Personnel Psychology, 50(3), 655-702.

Dipboye, R. L. (1982). Self-fulfilling prophecies in the selection-recruitment interview. Academy of Management Review, 7(4), 579-586.

Dougherty, T. W., Turban, D. B., & Callender, J. C. (1994). Confirming first impressions in the employment interview: A field study of interviewer behavior. Journal of applied psychology, 79(5), 659.

Hiscox, M. J., Oliver, T., Ridgway, M., Arcos-Holzinger, L., Warren, A., & Willis, A. (2017). Going Blind to see more Clearly: Unconscious Bias in Australian Public Service Shortlisting Processes.

Kuncel, N. R., & Hezlett, S. A. (2010). Fact and fiction in cognitive ability testing for admissions and hiring decisions. Current Directions in Psychological Science, 19(6), 339-345.

Le, H., Oh, I. S., Shaffer, J., & Schmidt, F. (2007). Implications of methodological advances for the practice of personnel selection: How practitioners benefit from meta-analysis. Academy of Management Perspectives, 21(3), 6-15.

Martin‐Lacroux, C. (2017). “Without the spelling errors I would have shortlisted her…”: The impact of spelling errors on recruiters’ choice during the personnel selection process. International Journal of Selection and Assessment, 25(3), 276-283.

Menkes, J. (2005). Hiring for smarts. harvard business review, 83(11), 100-9.

Moore, D. A. (2017). How to Improve the Accuracy and Reduce the Cost of Personnel Selection. California Management Review, 60(1), 8-17.

Paulhus, D. L., Westlake, B. G., Calvez, S. S., & Harms, P. D. (2013). Self‐presentation style in job interviews: The role of personality and culture. Journal of Applied Social Psychology, 43(10), 2042-2059.

Stoller, J. K., Goodall, A. H., & Bäker, A. (2016). Why The Best Hospitals Are Managed by Doctors. Harvard Business Review.

Dr Martin Dunn

REDACTED