Affirmative action is perhaps one of the more polarising phrases to enter our political lexicon in the last few years, which is darkly ironic given we have access to an increasingly vast body of data showing how affirmative action policies almost always end in abject failure. How is it that we still find ourselves debating the merits of gender quotas for politicians and enforced ethnic diversity amongst corporate leaders when it has been repeatedly proven that such policies do nothing to resolve whatever purported inequality or discrimination they claim to redress?

The formidable economist Thomas Sowell provided the answer to that question repeatedly and forcefully throughout his decorated career – because it sounds good and makes people feel as if they’re taking positive steps to ameliorate some of the injustices of the past. To what extent such injustices can be compensated for and the possibility of actually achieving such an outcome is largely ignored. Fortunately, we’ve long since moved past the era of entrenched systemic and legal discrimination.

It is then crucial to understand what is meant by the term affirmative action. Affirmative action refers to any specific, targeted or goal-oriented effort to reduce some measure of theoretical inequality between distinct groups. To give a simple example – we know that women are less likely to pursue careers in STEM disciplines. If such a law existed that prevented women from becoming mining engineers (but luckily it doesn’t, because we don’t live in Iran) removing such a law would not be considered affirmative action. Enforcing a quota or some sort of hiring target with the sole purpose of achieving a specific level of female representation in STEM roles would be affirmative action – and while this example may seem perfectly innocuous it is precisely this type of policy that is most pernicious, both socially and economically.

In practice, diversity quotas fail because they do not address the underlying causes of whatever creates outcome differences in the first instance. For a case in point, The University of Melbourne has set the seemingly simple goal of ensuring that the proportion of indigenous undergraduate students is in accordance with the total Australian indigenous population proportion – yet this goal doesn’t account for the fact that indigenous youth are far less likely to attend or finish secondary schooling.

Moreover, the social case against affirmative action is almost so strong as to seem ludicrous. Ethnic affirmative action policies have been tried in India, Malaysia, Sri Lanka and Nigeria and have resulted but nothing in civil unrest, race riots, political turmoil and violence. Funnily enough, most demographic cohorts are happy to coexist in peace until such time that a politician assures them that they’ve been the victim of gross historical injustice and that their fellow citizens were the perpetrators. What possible justification could there be for risking these awful consequences in Australia? Worse – what are the impacts if our public institutions continue to imply that such a dynamic of oppression and retribution should exist between the sexes? Despite the blindingly obvious flaws in supporting the politicisation of racial differences for the sake of “social cohesion”, this remains a favourite term of our (soon to be erstwhile) Racial Discrimination Commissioner, Tim Soutphommasane.

Attempts to enact affirmative action policies are associated with numerous direct and indirect financial costs – these range from the initial analysis of any perceived problem to the reduced efficiency of employees who were hired to fill a quota instead of on merit alone. Sadly, the Australian Public Service operates without the scrutiny applied to private business – unelected bureaucrats can endlessly dream up methods (and even construct new government agencies) to solve hypothetical social ills with only a cursory understanding of the financial implications. While it would be nearly impossible for an interested citizen to compile enough data to explore the extent of the issue (unless they had access to some of the second-hand furniture stores around Canberra) sometimes even the APS manages to provide some useful insight – albeit unintentionally.

In an attempt to expose the presumed gender discrimination and racial hiring bias in the APS the Behavioural Economics Team of the Australian Government ran an experiment within 14 government agencies to explore the effect of ‘de-identified’ resumes on employment decisions. The theory was that removing all information that alluded to gender or race would help managers overcome their own internal bias when hiring. From the report:

Introducing de-identification of applications… may have the unintended consequence of decreasing the number of female and minority candidates shortlisted for senior APS positions, setting back efforts to promote more diversity at the senior management levels in the public service.

You read that correctly – the average APS applicant enjoys an advantage during the hiring process by virtue of their genitalia or skin colour. To use the lingua franca of some affirmative action advocates, this is seriously problematic for many reasons, not least the fact that BETA seems to have shown that APS agencies are violating the Public Service Act (1999) which stipulates:

The APS is a career‑based public service that… makes decisions relating to engagement and promotion that are based on merit.

Ultimately, there can be no justification for the emergence of affirmative action in the APS. Any entity using taxpayer funds to operate is obligated to ensure that they are spent with maximal efficiency – the deliberate subversion of hiring practices in the APS is a precursor to exactly the sort of discrimination that is ostensibly targeted.

Private corporations are of course free to do as they please when endeavouring to mend social injustices – but as consumers and taxpayers, we can’t opt-out of the APS by taking our business elsewhere.

Illustration: BBC Television.

Got something to add? Join the discussion and comment below.