Friday, August 05, 2016

Harvard Business Review: Disastrous diversity training doesn’t work, drop it…

"You want the truth? You Can't handle the truth!" that great line delivered by Jack Nicholson in A Few Good men. It's what's happened in L&D for the last few decades. Is there any other profession that would put up with so much money spent to so little effect, resulting in so much failure?
In a brilliant article in the Harvard Business review this month, the most knowledgeable diversity researchers bar none, Dobbins and Kalev, explain, once more, that despite massive spends on diversity training, it quite simply doesn’t work. Let me repeat that – it doesn’t work. I’ve been writing about this for over ten years, presenting the evidence, talking about it at conferences but still I experience L&D as a profession that would rather just deliver courses, that don’t work, than think about solving the real business problem.
Merill Lynch paid out half a billion in fines over 15 years, no small sum. Yet the numbers of minority candidates and women senior managers have only increased marginally in most large companies. Dobbins and Kalev represent the evidence from 1000 studies in over 800 companies, over 30 years, to show that diversity training not only DOES NOT WORK, it is actually COUNTERPRODUCTIVE. The effect of compulsory courses is short-lived and can result in a backlash effect. Blaming and shaming doesn’t work and training should not be presented as ‘re-education’. You can choose to ignore this evidence, and ‘Keep becalmed and carry on’ or be taken seriously by senior management and do the right thing.
What doesn’t work?
Compulsory training
This doesn’t work and often produces negativity around its accusatory tone. Negative messages and implied threats don’t work. It often fails to change attitudes as it uses inappropriate presentation and exposition training techniques. The evidence is clear – avoid this as your main strategy. The evidence also suggests that it tends to particularly hurt women in the organisation.
Testing
Testing doesn’t work as the evidence shows that white men (managers’ friends) are often given a pass and the test results are interpreted inconsistently. Bottom line – managers fiddle the tests. As Dobbins and Kalev stress, it actually hurts minorities and women. Performance appraisals show similar results, with male managers showing bias in outcomes.
Grievance processes
The evidence suggests that this can produce conflict and a culture of accusation and complaint, as well as resulting in retaliation, negating what you’re trying to achieve.
What works?
Voluntary training
Voluntary training eliminates the forced and accusatory strain that often exists, when you are identified as needing such a course and also within the course, where you may be made to feel unnecessarily identified as a culprit.
Mentoring
This seems to have a positive, beneficial effect on actual diversity outcomes.
College recruitment programmes (minorities)
Targeted initiatives for minority groups in schools and colleges, work, as they are don’t plaster over the problem after the event but tackle the issue at root.
College recruitment (women)
Ditto.
Diversity programmes
These programmes, that tend to focus on management interventions and strategic progress, work better than crude compulsory courses.
Diversity manager
A Diversity manager will be able to identify a strategy and implement subtler approaches, as well as having the authority to effect change. This move has proven efficacy.
Conclusion

To genuinely build a constructively fair and meritocratic organisation that values everyone, regardless of race, gender and socio-economic background, most diversity strategies have to change. It needs a subtler, less compulsory and more managerial, strategic approach that eschews glib, compulsory courses and grievance procedures, in favour of more voluntary and subtle approaches, along with positive interventions pre-recruitment. In short, don’t berate with courses, test or use grievance procedures as a substitute for good management. Full HBR article here.

No comments: