Affirmative Action or Equal Opportunity?
This interactive demo shows a simplified model that captures the dynamics of a growing organization in which people are hired, promoted, terminated or retire.
The simulation can capture a number of phenomena that are observed in real companies. In this example we wanted to focus on biases that may exist during the hiring or promotion processes, as well as the impact of starting with an organization that is poorly balanced.
We wanted to explore the impact of introducing and removing biases. In particular, the assumption is often made that, in order to improve diversity or gender balance, it is sufficient to remove biases in hiring and promotions, and to focus on expanding the pipeline of incoming talent.
To explore these issues, we tested two simple scenarios. First, if the company starts out being well balanced (50-50 gender split) at all levels, and biases are introduced in the hiring or promotion, how much impact will be felt in five years? To test this, set the initial imbalance to "None," set the Promotion Bias Toward Men at 20%, and hit the Run button. Although each run will have different results because of randomness, in general you will see that after five years the company has that all-too-familiar "pyramid" shape, with men dominating the upper echelons.
Next, we wanted to know whether, starting with an imbalanced company, it is sufficient to remove biases in order to restore balance. In other words, is Equal Opportunity enough? The short answer is a resounding "no." To see this, try setting the initial imbalance to "Low" and make sure both bias sliders are set in the middle (no bias). After five years, the company will look better, but still the upper levels will be heavily skewed toward men.
Our Forbes blog about this simulation provides more details and includes results obtained by running hundreds of runs to get meaningful "average" behaviors. However, we encourage you to play with this simulation to test various scenarios. For instance, try setting a Low or High imbalance, set the promotion bias to zero (midpoint), and set the hiring bias to -20%, which means that the company is actively hiring more women than men. Run for five years and you will see a configuration that is all too familiar in many industries, with more women than men at lower levels, while the men continue to maintain a grip on the upper levels.
Please note that, just as in real life, there is randomness in the way the model works. Hence each time you run the simulation you will get slightly different results, and occasionally you may get "outlier" results. For instance, even when you start with a balanced organization and have no biases, you may end up with some strong imbalances after five years.
Finally, this is just a simple example to give you an idea of the power of using computer simulations to understand the link between diversity (in this case, gender diversity) and performance. If you are interested in learning more, or you want to see how these principles can be used to help you explore much more complex scenarios for your organization, please contact us.
The material on this page and the simulation are copyright of Aleria. This content may not be modified or reproduced without our prior permission.