(Can) Fight the System

Hannah Rosin, to promote her new book The End of Men, wrote an article on Slate this past week claiming that patriarchy in the United States is dead. ‘Patriarchy’ has a variety of different definitions, but in this context I believe Rosin is using the term the way feminists typically use it, to mean the dominance of men and the consequent subjugation of women. Her article has caused quite a stir. I want to focus on a particular aspect of her argument: the failure of systems thinking that it represents. It feels a little unfair picking on Rosin, since she is by no means the only person demonstrating this lack of systemic insight in a major media outlet. However, because her article touches on the main subject area of this blog, I think it’s useful to use it as an example of how to change the thinking and dialog around the topic of gender equality.

I’m a systems scientist, working in environmental systems with both human and natural components.  A large part of my job involves building models of systems—on paper or on the computer—in order to help people who make decisions to consider both the long-term consequences of their actions and the unintended outcomes of those actions. The central concept in systems theory, which has been applied in a wide variety of fields, from business management to the health sciences to ecology, is the concept of emergence. Simply put, emergence means that you cannot understand the behavior of a system by looking at its component parts in isolation. One example is that of baking a cake—if you tasted the flour, the eggs, and the sugar separately you would still have no idea what the final cake would taste like. The taste of the cake is a property of the interactions among all of the ingredients.

Similarly, I don’t think you can find patriarchy by looking at one’s interactions with one’s husband, one’s boss, one’s male colleagues/classmates in isolation, as Hannah Rosin seems to claim. John Sterman of MIT, one of the most prominent systems scientists working today, puts it this way: systems ‘have no boss’. In this context, that means there is no patriarchal cabal tucked away in a wood-paneled conference room smoking cigars and sipping single-malt scotch while cackling, “BWAH HA HA! Our efforts to foil the progress of women-folk are PROCEDING AS PLANNED!” Patriarchy, like racism, environmental degradation, obesity, and many other systemic problems, is not any individual’s ‘fault’. On the other hand, all of us are participating in the system and responding to its subtle pressures, so all of us are in some sense contributing to these problems. This is why we need systems analysis—so that we can act more deliberately and thoughtfully by considering the full set of information about the consequences of our decisions.

Here are some reasons I, personally, believe the patriarchy is not ‘over’ in the United States (I think we can all agree that womens’ rights in many other parts of the world have a long way to go)

1. The wage gap—women still earn $0.77 for every dollar a man earns, and this gap has not budged in a decade (more on this below).
2. The appalling rates of sexual assault and domestic violence in this country, particularly among poor, rural women. Plus, the amount of victim-blaming that still goes along with these crimes.
3. Our maternal mortality rates are among the worst in the developed world—again, particularly for poor women and for women of color.
4. Women make up the majority of those living in poverty in the U.S.
5. Women are still very under-represented in leadership positions in politics, academia, business, etc., as we have discussed in this blog before.

I could go on with more examples. In fact, this entire entry could probably be made up of examples. However, I believe it would be more instructive to take one of these and demonstrate some ‘systems thinking’ around it.

There are several causes of the wage gap, but let’s pull out just one: motherhood. I want to demonstrate a reinforcing feedback loop that can serve to keep the gender pay gap intact, using academia as a case study. A reinforcing feedback loop can be seen as a ‘vicious cycle’—because initial conditions are the way they are, a certain outcome tends to happen, which in turn reinforces the initial conditions.

Say a young married couple get jobs as assistant professors at the same time at a research institution. Even if they are in comparable fields (say, both within the sciences), the husband is likely to be paid more. Then, the couple decides to have a baby. The university is generous by U.S. standards—they offer twelve weeks of paid maternity leave. But the couple decides this is not enough; they quite reasonably would like to have six months of time at home with their baby. Because the husband’s salary is greater, they both agree it makes more sense for the wife to take an unpaid leave. Consequently, her research productivity is reduced. Her time to tenure is lengthened. Her husband works long hours to make tenure—it makes more sense now for him to push to do so, since his salary will be even larger once he gains tenure. The wife picks up more of the childcare and house-care responsibilities as a consequence. Slowly, she finds herself slipping behind. She decides to convert her position to a non-tenure track research associate, or to an adjunct position. Her take-home pay is further reduced, but her hours are now more flexible. The department that hired her originally feels that they didn’t get their money’s worth, since they wanted someone with high research productivity to bring in federal grant money. The next time they hire a married woman, they will keep this in mind with the initial salary they offer her. So, initial wage gap à rate of women dropping out of tenure track à initial wage gap. This is a feedback loop. Who is to blame here? The husband? The wife? The university? The department chair? None? All? In order to resolve this problem, we have to understand the mechanism, and then we have to identify levers of change. In this case, it might be a university policy on gender-equal pay. Or, it might be low-cost, onsite childcare. The efficacy of both of these policies could be tested by using a systems model. The moral of the story is: don’t look for a person to blame. Look for the mechanisms underlying the system that keep certain outcomes in place. Trust a systems scientist—this is a far more productive exercise.