If you’re analyzing data related to people, odds are good that you’ll have to include some demographic data. Understanding the age, ethnicity, gender, and status of the population you’re studying is important. You need to think carefully about how to use that data to represent the lived experiences of a diverse group of humans.
The We All Count position is that using an equity perspective when using data not only makes it more fair, but makes it much more robust, and usually more accurate. To ensure equity in your analysis, it’s critical that you use data to reflect the fact that a person’s experiences are based on multiple dimensions or identities. Different combinations of demographics create different types of people.
Let’s look at it another way. Understanding how outcomes vary in your analysis between men and women is important. But it’s also crucial that you acknowledge that different types of women have very different experiences. People’s social identities are based on combinations of gender, ethnicity, ability, age, class, mobility status, and more. Identities can’t be reduced to a single dimension. I’m not just a woman. Or just a Canadian. Or just a university graduate.
From an equity point of view, it’s essential to find a way to use our data and analysis to reflect this. Otherwise, we risk rendering this diversity invisible by looking at individuals only one dichotomy at a time.
DO I NEED TO USE DEMOGRAPHIC DATA?
“But the entire purpose of my research is to establish some evidence on how a person’s outcomes in my project are different based on characteristic X!”
That’s ok.
It is entirely possible to build a model that accounts for a wide variety of different combinations of lived experiences and identities and then report out those results in more or less aggregated ways if that’s what you need.
UNDERSTANDING INTERSECTIONALITY
“Intersectionality is a metaphor for understanding the ways that multiple forms of inequality or disadvantage sometimes compound themselves and create obstacles that often are not understood among conventional ways of thinking.” – Kimberlé Crenshaw
If I have the relevant data on each of the demographic categories in my project, can’t I just easily combine them? I can clearly see the impact of being male, or the impact of being Hispanic, or the impact of being in a mid-income category, why would I need some specific intersectional profile that’s different than the sum of those factors?
It’s easier to illustrate intersectionality than it is to explain it. The term was coined by Columbia Law and UCLA Law professor Kimberlé Crenshaw. And she does a great job of laying it out in her TED talk. Here is a clear, concise explanation of what it really means in one person’s lived experience:
Remember how I talked before about the issues with viewing individuals based on a single facet of their demographic data?
We’ve all heard about the gender pay gap. In this instance, women are the marginalized group; women in the US earn, on average, 72 cents for every dollar their male counterparts earn. But it might surprise you to know that the problem isn’t just that women as a group earn less than men.
Women who also belong to ethnic minorities (black, Latina, and indigenous populations) make even less than white women. In fact, the Economic Policy Institute reports that the wage gap between white women and women of colour is growing faster than any other wage inequality.
CUMULATIVE MARGINALIZATION
Let’s look at a study on political participation. The goal was to understand how gender, ethnicity, and social class intersect to influence an individual’s participation in the political system by means of protest. Let’s focus on the demographic data of one of the study participants.
They are:
- Female
- A member of an ethnic minority
- Representative of a lower social class
How will this subject’s demographic data affect their participation in the political process?
Intersectional identities influence participation in politics. In general, a person is less likely to participate if they belong to a marginalized group. If we look at how likely a man is to get politically involved compared to how likely a woman is to get politically involved, we see that the data shows a woman is less likely. The same is true if we compare a person from a lower social class to a higher social class. This is an additive approach – looking at one characteristic at a time.
However, in reality, no one is living one of their characteristics at a time. My lived experience is a simultaneous combination of all my characteristics. If we want to use data to more accurately reflect my lived experience, we need to use it combined as well. This is a multiplicative approach. This allows us to see how different characteristics are interacting – or intersecting. When we do this, the data shows us that the cumulative marginalization is much stronger than the marginalization of just one group at a time. When individuals are members of more than one marginalized group, the effect can be cumulative.
In our example on political participation, the analyst takes a multiplicative approach to intersectionality. They suggest that the relationship between an individual and the attitude or behaviour they hold is conditional on identities intersecting.
In the political participation example, researchers examined four interactions:
- Gender and class
- Gender and ethnicity
- Ethnicity and class
- Gender, ethnicity, and class
THE EFFECTS OF INTERSECTIONALITY
The chart above demonstrates the effects of cumulative marginalization. The higher order of intersection (the number of marginalized identities a person holds) the lower their probability of getting politically involved.
From the chart we see that a woman has a probability of getting involved around 17%. A woman from the lowest social class has a probability of getting involved of 9%. However, a woman from the lowest social class and also belonging to the marginalized ethnic group has a probability of getting involved of only 7%. This is using the data multiplicatively instead of additively to highlight the lived experience of cumulative marginalization.
If we only look only at each facet of our study participants in isolation (men vs. women, ethnic majority vs. minority, or upper vs. lower class), the results look very different. And more significantly, looking one facet at a time renders the plight of those individuals belonging to multiple marginalized groups almost entirely invisible. With an emphasis on using data for equity, it’s important to look at humans intersectionally.
Using demographic data in your research isn’t always easy — but it’s crucial to ensure your analysis is equitable. Want to learn more about incorporating equity into your data? Come learn with us at We All Count.