Gender segregated buses for Jakarta's Busway?

In one of my first posts, I talked about the reasons that societies sometimes choose to segregate vehicles, or sections of vehicles by gender, and the costs associated with doing so.

Coconuts Jakarta is reporting that Transjakarta, the operator of Jakarta's busway service, is considering introducing female-only buses to combat sexual harassment. The comment was made by the President Director of Transjakarta in response to a recent incident of sexual harassment that occurred on a public footbridge after both patrons had been riding a Transjakarta bus.

In my abovelinked post, I wrote about how gender segregation on public transport decreases the efficiency of the fleet, and imposes some other more difficult-to-value costs on society. In the case of completely gender segregated buses, as opposed to gender segregated sections of buses, the impacts on efficiency will be extreme.

When you have impacts on efficiency, one of two things happen: passenger wait times increase, or your fleet size needs to increase. At present, the fleet in Jakarta is well below what is needed,  and they're currently scaling up as fast as they can but it will take a few years until the fleet size is sufficient, no matter what the gender segregation policy. So, in Jakarta, an efficiency impact will flow pretty much entirely  through to wait times for at least the next few years.

If this is being seriously proposed, the biggest questions for me would be:

• What proportion of buses would be women-only?
• Would the gender-segregated sections in existing buses be preserved?

Currently, around one third of bus space is segregated for women. This means that, if you don't maintain the gender segregated sections in existing buses, to maintain the total women dedicated capacity would mean every third bus would need to be women-only. This would mean that, on average*, if you're a women that wants to ride in a dedicated section, you'd need to wait 3 times as long to get that dedicated section. You're also increasing wait times for men by 1.5 times**, because now they'll only get two buses that they can ride on in the time they used to get three. 

If you do maintain the gender-segregated sections, and offer a smaller fleet of women-dedicated buses, those buses will be even less frequent, so wait times for women wanting a completely segregated bus will be more than 3 times the current wait times.

If forced to wait for 3 (or more) times longer than they currently do, would women still choose to ride the Busway, or would they switch to other forms of transport, or elect not to travel? 

This particular case of sexual harassment would have been prevented had the victim in question been travelling on a gender segregated bus, as the harasser followed her off the bus. However, had that particular woman been on a gender segregated bus, the harasser may have selected another woman on the mixed bus, or just some other woman on his walk home that wasn't riding the bus at all.

The costs this policy would impose would be significant, and the largest burden would be borne by women. If it is seriously being considered, Transjakarta and the city government of Jakarta should first publish data on sexual harassment rates on the Busway, their estimates of how many incidents would be avoided by this policy intervention, and the expected impacts on travel times. Following publication of this analysis, they should undertake a survey of the Busway's passengers (disaggregated by gender) to see if the change would be welcome. 

When an incident occurs, the natural instinct for a lot of people is to demand change to make sure it never happens again. But, in situations like this, the rush to show concrete action can result in policy outcomes that can be counterproductive. Sometimes the cure is worse than the disease. 

---

*Ignoring peaking effects for simplicity.

**Ditto

Gender and infrastructure

In my previous post I delved into the thorny issue of gender and infrastructure. It's something I have been thinking about a lot recently, both personally and professionally. 

It's the kind of area can sound a little odd. I mean, let's be honest here, the gender impacts of the alignment of a power transmission line aren't really immediately obvious. That said, in certain cases, like public transport policy, there can be some really significant positive or negative gender impacts.

In my experience, gender in infrastructure has been viewed through a sort of a safeguards lens, where you plan your project, then check afterwards if there are any negative gender impacts. If so, depending on how much they cost, you might do something about them. I'm interested in whether that is in fact the best way to do things, or if there's some other way. And, if there is some other way, how we can build systems that incorporate these issues from the ground up.

In Indonesia, pretty much the only people I know that are looking at the practice of this sort of thing are the Indonesia Infrastructure Initiative (IndII), an Australian government-funded program*. They have some great materials on gender and disability and are actively incorporate these issues into their program design.

I have also been lucky enough to have a chance to discuss this topic with Dr. Lisa Cameron of Monash University in Melbourne, whose work on Indonesia I have loved for a long time and highly recommend. I'm hoping she sees fit to add to the academic literature on this sort of thing...

Outside of this, and what I have found with a bit of a google, I'm not that aware of many good resources on the topic. If anyone knows of any good research or interesting researchers looking at this field, please let me know in the comments.

---

*Full disclosure: I worked for AusAID on the design of IndII (I even came up with the name), and collaborate a lot with IndII on a day to day basis in my day job.

Do we need gender-segregated public transport?

It's not uncommon for societies to make the decision that providing women's only spaces on public transport is a good thing. I'm most familiar with it on Jakarta's Busways, but I remember seeing it on commuter trains in Mumbai, and during the rush hour in Tokyo. It has even been discussed in Australia and the UK at various points.

The most common justification provided by proponents of the policy is that it creates a safe space for women so that they can be free of harassment. This can be a really important reform in places where a significant population of women would choose not to use public transport because they feel unsafe. Women that feel unsafe will either be pushed towards private transport at higher personal cost (taxi fares, or running costs, depreciation for a private car) and social cost (traffic), or choose not to travel at all. Choosing not to travel  can have big negative impacts on women's workforce participation, or if they wouldn't otherwise work, mean that they have to shop at the expensive minimart near their house rather than the cheaper market further away, or force their husband/son/male family member/friend to take more time out of their day to run errands.

So, that can be a big positive, but in policy, we need to weigh the positives against the negatives and the potential positives of all other options to see what the right decision is.

One thing I personally hate about gender segregation is that it puts the onus on women to act to not be sexually harassed. This sort of thing can give rise to statements like "well, if she didn't want to be harassed, she should have been sitting in the women's section." As the harasser is the being anti-social, ideally they would be the one required to go out of their way to stop this happening. Sadly, that's not practical. 

Gender segregation also decreases the efficiency of the public transport system, by decreasing the fungibility of the seating. If you've got capacity for 50 people on a bus, how do you split it? If you ridership is 60% men and 40% women, let's assume for simplicity that you allocate your seating 30:20 men:women. Through random chance and due to differing travel patterns, you're sometimes going to have times when one section will be full and the other section will have capacity, meaning that, at the margin, you need a bigger fleet (or longer waiting times) to handle peak demand. You can (as all services I have seen do) make the "women's" section a "mixed" section, and adjust the split to account for the fact that some women will be indifferent or may choose to sit in the mixed section, but you'll still run the risk of the mixed section reaching capacity and men waiting longer, or the women's section reaching capacity and women that don't feel comfortable riding in the mixed section waiting longer.

So, how do we balance the positives and negatives? Is gender segregation a good thing? 

To me, the clear answer is: it depends. And, no, I'm not being facetious...

I have never been to Saudi Arabia, but I understand that women aren't allowed to ride the most common forms of intra-city buses there, and aren't allowed to drive. So if women in Saudi Arabia want to get around "by themselves", they have to hire private drivers. In a context like that, I could imagine that women-only sections or whole buses could be a great thing for their society. But, I think I would be a bit disappointed if we had to do this in Melbourne, Australia (where I'm from).

I hate the term "international best practice." It is used a lot in development policy (I'll write a bit about the term and its problems more in a future post), and it sort of implies that there's a single solution that is going to be the "best" solution in every situation in every society. Or, if not that, it implies that some "expert" (usually and "international expert" in development) can come in, look at some some numbers and design some solution that is the "best" outcome for that society. This is almost never the case.

I don't think quantitative analysis is useless in these sorts of discussions--it's important to know how many women ride the buses, how long waits are likely to be, see how many women ride in the mixed sections, and how these things change over time--but, to me, the quantitative analysis is an input into the qualitative discussion about what value the society puts on the positives and the negatives.

Someone always gets screwed whenever you bring in a few policy (or, at least, benefits less than someone else). Gender segregation directly benefits a proportion of women that use, or would like to use public transport, and directly harms all men that use public transport (there may be second order-benefits, but they definitely bear a cost). So how do you balance the costs and benefits not just to society, but between individuals? There's not some best practice formula that says women's time is worth this much and men's time is worth that much and this is how you trade them off. This is where politics comes in. 

Politics is often a dirty word in policy. People say "oh, the other side were playing politics" with a sneer. But, to me, in a well-functioning democracy, we vote for people that we think have some approximation of our values. The politicians, for better or for worse, are as close as we get to a representative sample of the democratic unit. My job as an analyst is to get as much data together as possible about who bears the costs and gets the benefits, and how big these costs and benefits are, then the politicians get to decide how to allocate them, based on their understanding of the values of their electorate. If they do it wrong too often, they (hopefully) get voted out and someone else gets to have a go. I have my own values, and they do affect how I present my analysis, but at the end of the day, it's the politicians that have to face up to their citizens, not me.

---

Quantitative geek that I am, I love seeing data like the graph in my header showing what improvements men and women want to see in transportation in Dhaka, Bangladesh.  I find it fascinating that there's such a big gender gap in the proportions of men and women that want separate buses for women. I do wonder how big gaps like that impact upon how controversial policies end up being. I have no idea what the best policy solution is here, but that gap indicates that you'll probably have to do a lot of work building consensus around a solution.