Speaker diversity

I was really pleased to have a session accepted for the PASS Data Community Summit in Seattle this coming November. The acceptance email said that 516 individuals had submitted 863 sessions, so it was a real honour to have been selected. I’ve attended many PASS Summit events over the years, including the virtual ones in 2020 and 2021. My first was in 2010 when I delivered two sessions plus a lightning talk, and I delivered another two sessions plus a lightning talk and added a pre-conference seminar in 2011 – the year when I also got on stage during the closing keynote. I served as a PASS Board member for 2.5 years about ten years ago, and continue to tell people that the PASS Summit is worth attending.

The world seems to be on-edge at the moment. And I don’t really mean “at the moment”, because there has been a progression, and I don’t think we’re going back. I should say “now”. We have moved our consciousness to be more inclusive and more equitable. We understand better (not completely, but better than before) that people’s circumstances are often not just a result of their own actions, but because of things outside their influence, and because of the way society handles them. Privileged people like me – tall white English-speaking able-bodied cis-gendered straight men have an easier time of life. We play the game on the ‘easy’ setting. Not by choice, but just because society affords that to us. I’d like to think that one day everyone will be able to enjoy the same privilege. I doubt that will happen any time soon. The world is “on-edge” because addressing diversity seems to be strangely contentious. Many people who have privilege don’t want to give it up. Many others are happy to go along with the status quo (which is almost the same stance as the privileged trying to hold onto privilege). But the biggest opportunity for change is for privileged people (like me) to try to do something. (And I know I don’t always get it right. I’m trying. I want to keep improving in this.)

So when the speaker email came out, I replied by asking how the diversity was looking. I also accepted the invitation to speak. I was a little surprised by some of the things in the reply. And I want to make it clear that I am not trying to attack Red Gate (who have produced the event since 2021) for the numbers that have come through. They have explained their desire for diversity and I admire that. This post is to provide suggestions on where I think it falls down, not to criticise them for a lack of trying.

They told me:

I don’t know how many speakers were selected. I know there was a pool of 516. And I know from past events that they will have around 200 sessions, with many speakers giving two sessions. And there will be plenty from Microsoft employees that haven’t been decided yet. So let’s assume there have been 130 speakers chosen. Of those, 62% provided answers to the diversity questions. I don’t know whether the people who didn’t answer are more diverse or less diverse, so I’m going to assume the same. (I happen to think it’s probably less diverse, but obviously I can’t be sure.)

22% of 130 is about less than 30. That includes women and anyone else who specified they were non-male. Less than 30 women, and over 100 men. 78% men. Nearly four in every five speakers is male.

And I could say similar about the 82% white people, given 18% reported non-white race or ethnicity.

43% of the speakers belong to at least one of the categories listed, meaning 57% are white able-bodied men from North America or Europe. And I’m in the 43%, because I live in Australia.

To me, this doesn’t feel like diversity goals have been met. But again, this is not about attacking Red Gate. It’s good that they want to address this at all.

The people at Red Gate wrote an open letter this week to the community, in which they addressed the diversity. They describe four methods they are using to promote diversity.

• Mentorship & encouragement
• Speaker invitations
• Anonymous submission review process
• Engaging the power of the community towards continuous improvement

Three of these I’m fine with, and again, I really love that they are making concerted efforts to develop diversity. But “anonymous submission review process” is an interesting one, and I wrote some tweets about this. After all, for all the community engagement and mentoring / encouragement, the submission review process is still the gateway that needs to be passed through to become a speaker, unless someone has bypassed it by invitation (and if there are lots of people invited, that means the ones chosen by the review process must be even worse than 22% non-male).

Some of those tweets (but there were quite a lot)…

I don’t think completely blind submission evaluation helps diversity. I think it’s reasonable to pull the names, but you should leave the diversity information on there. Otherwise, you’re bound to favour those who are better-practised at writing submissions, ie the privileged.

— Rob Farley (@rob_farley) June 24, 2022

And I think they do have a desire to do better. Impartiality in speaker selection isn’t it though. That doesn’t work for developing diversity. We need to be choosy about who we put in, or else the under-privileged will be under-represented.

— Rob Farley (@rob_farley) June 24, 2022

Selected sessions without knowing who has submitted them was an initiative introduced some years back because of a fear that the same names were being chosen year after year. By removing the names from submissions, the program committee was able to select the sessions that they thought were the best without being influenced to pick the big names. Let me emphasise – this was an initiative to try to help address issues. It’s well-intended. That’s a good thing.

But I think it just means that the sessions that get chosen are the ones created by the people who are well-practised at creating sessions. The same names year after year.

And strangely, the criteria that is being used to select the sessions – the abstracts – aren’t generally used to advertise the conference or even consulted much by attendees at a conference, and they could be changed after the speakers have been selected. I hear about abstracts being dismissed because of spelling or grammatical errors (even because of people not using a ‘z’ in ‘optimise’, which is purely a translation error), which inadvertently favours people for whom American-English is their first language.

I think (and I’ve told this to people involved in discussions with Red Gate) that the the speaker pool should be chosen almost independently of the session abstracts. The specialisation areas of speakers could be a consideration, but also their speaking experience, their ability to draw a crowd, their expertise, and their diversity – satisfying the commercial side of the event with the need to reflect the diverse nature of the community. A mix of new blood and old blood can be selected, while ensuring strong coverage of specialisation and a balanced group of individuals. After creating a draft pool of speakers, the abstracts can be considered, and changes can still be made. Topics that aren’t covered could result in changes to the selected speakers, but could also be addressed by asking the selected speakers if they could create a session on that topic. If no one has submitted a session about database corruption, and the organisers want this covered, the speakers who often talk about this could be approached to see if they wouldn’t mind – regardless of whether or not they submitted a talk on that. That should already be the case, if topics are not covered by the 863 (this year) sessions submitted.

And despite tweeting that it’s reasonable to remove the names, I don’t even think it’s a good idea – reasonable, but not my preferred option. I think it’s better to see the names, so that the selection committee can consciously include or exclude a person who has been a speaker every year for the last five conferences. Make the decisions on an individual basis. This also helps address cases where people are known to be contentious and should maybe excluded, or at least flagged for further discussion.

Anonymous means equality, right? It’s that thing of refusing to do favourites. But we end up with the age-old image (and I don’t know who created it originally) comparing equality with equity, showing people who want to see over a fence to watch a baseball game. By trying to do ‘fair’, the people who get selected are the ones who are already ahead, and the diversity drops off even more.

I simply don’t think speaker selection should be anonymous, because it makes selection harder for many people.

I realise that hand-selecting people might not seem fair. I appreciate that. But a selection committee can still use a variety of metrics in coming up with their list. If they need big-name tech-celeb types, they can pick some of those. If they want a good portion of people who are in multiple diversity categories (intersectionality), they can pick some of those. It doesn’t make it less fair. If you want to grow your profile or become a better presenter, so that you are more likely to get selected for a big conference, do that! Speak at events that will take you. Write a blog. Answer questions. Become known for your expertise. It’s not less fair to use this as one of the criteria for selection. I argue it would be less fair to ignore someone’s profile.

How can you make sure the voices from different demographics are heard if people’s diversity is completely ignored when selecting speakers through the normal channels? I don’t think you can.

I’m somewhat embarrassed to be writing about all this as a white able-bodied male. Remember, I count in a diverse category because I live in Australia. On a speaker page, I look like I have the same lack of diversity as the 57% who aren’t in a diverse category.

And so I have written to the organisers to ask that my speaking spot be given to someone in a more diverse demographic than myself. I apologise to anyone who was looking forward to seeing me in Seattle. I was looking forward to going too, but feel it would be awkward now.

I have also offered to provide mentoring to help. I already mentor some people in their journeys to be better presenters (mostly from diverse demographics), and am willing to take on more of this. I had submitted a pre-conference seminar about presenting – it didn’t get selected, but I am more than willing to help use that content to help people step up.

I don’t know who will be given my spot, and I’m not going to ask. I will trust Red Gate to choose someone appropriate. It’s none of my business, to be honest. I know I should not have accepted the spot and then be pulling out so soon afterwards. I hope they have time to react.

I apologise. But I hope that by giving up my spot, the diversity may increase. I don’t need to be a speaker at the PASS Summit this year. I will try to speak again in the future. I will continue to cheer for community efforts. But hopefully my spot can be taken from someone who is not a white man.

@robfarley.com@bluesky (previously @rob_farley@twitter)