Hiring discrimination - critique objectively

We refer to Koon Yew Yin’s response to our research on racial discrimination in hiring.

We regret that Koon could not attend our Nov 2 seminar where we presented our framework and findings. He could have gained a proper understanding of our research and appreciation of the study’s contributions and limitations. We welcome his response and his concern and suggestions for further study, but must correct his misunderstandings and misguided criticisms.

We make our main objective firm and clear. We set out to answer one central question, encapsulated in the title, “Does race matter in getting an interview?” Our subtitle notes that we conduct this study of hiring discrimination through a field experiment.

Koon quotes the title of our presentation in full. Yet oddly, and unfortunately, his criticisms basically revolved around the fact that our study does not provide substantial answers to the “why” questions. But that is not our objective.

Koon makes no substantive criticism of our method or findings, except to note that it is “not rocket science”. He dismisses our study for not being the study he would prefer we would do, and makes an overblown remark on our letter to The Malaysian Insider.

Let’s be clear: this study is sufficiently complete. It is far from perfect, but we are clear about our focus and limitations, and we provide empirically substantiated answers to the questions we pose.

Let’s also be clear that this study does not answer all the questions, those burning questions that we too will like to see answered, revolving around racial discrimination.

Shouldn’t we first assess the extent and patterns of discrimination before we venture to figure out why? Moreover, don’t we need to study where and how discrimination occurs in order to frame our “why” questions? Good research, which we strive to do, must be focused in its questions, methods and answers, and ought to raise questions for further research – in other subsequent studies.

How might we study such a delicate and difficult thing as racial discrimination? Koon recommends surveying employers. Of course researchers can pose questions to fresh graduates about of their job application experience and ask employers about their selection process. Such studies can be informative – especially to inform why discrimination happens – and it will be great if someone does them.

But the survey method faces a fundamental problem: respondents will have their own highly subjective perceptions of what constitutes discrimination, and they are likely to be biased or exaggerated in their responses, say, if graduates feel aggrieved over perceived discrimination, or if employers do not want to reveal their true motivations.

So we elected a method that we believe is more credible. Field experiments on discrimination, validated in scholarly studies around the world, observe real decisions made by employers on who they call for interview, not their declared intentions.

This is a far better way to gauge the existence and extent of discrimination. It is also a laborious process. Our study took seven months to produce and send 3012 resumes to 753 job openings, and three months to obtain data on the companies in our sample. We created four fictitious resumes for every job ad: one each representing above average Malay, above average Chinese, below average Malay and below average Chinese.

These categories were based on cumulative grade point average (CGPA), which is the most important resume feature in applicants’ interview prospects in the engineering and accounting jobs we applied to. We then record callbacks for interview, and observe whether resumes of one race are significantly more likely to be called after controlling for quality, which within our framework constitutes racial discrimination.

Through this method we can more objectively investigate to what extent quality matters, and to what extent race matters. We find that race matters a lot more than applicant quality as captured in resumes. We must clarify here that when we say we “control for quality”, what that practically means is that almost the same number of Malay and Chinese applicants have a CGPA above 3.6, which would be considered top of the class.

It is important to draw attention to this, because a common response, also articulated by Koon, refers to scenarios or experiences of Malay applicants on average holding lower academic achievement and therefore not qualifying for interview. Our data show that top of the class Malays, even graduating from the more reputable local universities, are substantially less likely than average Chinese graduates to be called for interview.

We also find interesting relationships between characteristics of applicants, of job ads and of employers, and the chances of getting called for interview.

Koon seems to credit us for one of these interesting results – the lesser likelihood for a Malay applicant to a Malay-controlled engineering company to get called for interview, compared to a Malay applicant to a Chinese-controlled engineering company. We also find that Malay applicants who state that they can speak and write Chinese have a better chance that those who do not. These are among our complex and nuanced findings that require further investigation through other research projects. Our study raises questions that most of us would not have asked, and this is a useful contribution.

But to Koon, our study is incomplete or difficult to receive because we have not obtained data in this study to immediately answer why we obtain our results. We can understand his eagerness to get to the bottom of things, but in the process he makes some hasty remarks.

To quote him: “What the two researchers have done is to allege the factor of racial discrimination without even interviewing the employers in their sample and examining deeper the reasons! Now what kind of research is this?”

This is careful, methodical and conscientious research. We do not “allege” racial discrimination – we seek and find evidence of it. Remember, we do so covertly; everything goes out the window if employers know we are doing this research. Therefore impartiality and confidentiality are paramount, and it would be an unconscionable breach of ethics if we were to contact and try to interview these companies. Further research will have to start over with another sample.

Apparently, the editors of The Malaysian Insider were also over-eager to assign blame for the gap between Malay and Chinese chances for getting interviewed. Our letter to them, which Koon cites, took them to task for the headline, “Malaysian employers practice bigotry, study shows”, which grossly misrepresented our work.

Koon claims we are “making a mountain out of a molehill” in making a fuss and asking for an apology. He is wrong on both counts. First, it is not a mountain. We never asked for an apology. It’s not about us, but the study and the public discourse that springs from it.

We expressed disappointment about the headline, and noted that the Bahasa translation of the article was reasonable and not carried with reckless sensationalism. We demanded that the headline be amended. They have defended their prerogatives, and we have moved on.

Second, it is not a molehill. Imagine if you were seriously misquoted in a loudly splashed headline. That’s effectively what happened, so we protested. Some people will have an absolutely wrong impression of our study lodged in their minds – that’s the power of headlines.

But we’ve been more concerned about subsequent comments and debates. We have spent much, much more time crafting responses like this than the few minutes it took to write the letter to the Malaysian Insider.

We feel the need to reiterate Koon, and all readers, that our study raised particular questions and offered answers to those questions. After extensive data collection, we find evidence of racial discrimination and demonstration that the problem is not simple and straightforward. We too would like to know why, but for now cannot offer substantial answers backed by empirical evidence.

Koon’s makes welcome and valuable suggestions for further research. Of course, employers’ perspectives and determinants of interview selection and hiring decisions should be examined. Public sector recruitment, which we tried to cover in this study but were unable to, warrants scrutiny.

But the challenge of filling in the many gaps in our understanding of this thorny, slippery and vital subject is not just for us to take up. It is for the academic community at large, for government, for NGOs, for thinktanks – indeed, for all of society.


The authors are from Universiti Malaya and Unversiti Kebangsaan Malaysia who organised a joint seminar titled 'Does race matter in getting an interview? A field experiment of hiring discrimination in Peninsular Malaysia'.