Should we have term limits for the AB (and possibly TAG)?

[tantek]

As @michaelchampion offhandedly referenced in #31 (comment), and given that we are moving more actual AB work to public CGs (like the Process CG) where people can keep doing work as they please, it is perhaps time to consider introducing term limits for AB participation, and possibly the TAG. We can do this incrementally with just for the AB at first (e.g. for Process 2019), and assuming that works well, introduce it similarly for the TAG (which may have additional complexities like should term limits apply to appointments also? (likely should))

[dwsinger]

random thoughts: I have never really understood term limits for posts like this; is there such a strong incumbent-advantage that better-qualified people are standing but failing to get elected (the problem that term limits purport to solve in public elections)? surveys I have seen also suggest that people want term limits and want to re-elect their own representative, which suggests that they want to use term limits to get rid of other peoples' reps, ahem.

I note however, that the TAG appointees tend to be the same year after year, which suggests that the appointment process isn't being used to get balance, but to retain good people. I have previously wondered why we don't require that appointees in year N stand for election in year N+1...the Director could still appoint them if they fail to get elected, but it gives valuable feedback on how the community feels about someone, and if there is an 'incumbent advantage', then the Director can find someone brilliant but not widely known, appoint in year N, and then they become known as brilliant and stand a much better chance of getting elected in year N+1.

[michaelchampion]

There's a dilemma -- the most QUALIFIED people on both the AB and TAG are almost certainly those who've learned how the system works and what the fixable and unfixable problems are. But both groups need to be stimulated/challenged to look at old problems from fresh perspectives and new problems without old baggage.

A related problem is that there are highly qualified people willing to contribute to the work of the TAG and AB but don't get elected/re-elected due to the various quirks of any election system. A process that creates more vacant seats to encourage people to run, and increases the pool of people with AB and TAG experience that they can apply in the community outside the formal advisory groups.

Not to beat my favorite dead horse or anything ;-) but term limits could help mitigate the downside of recent election changes that make it easier for single-issue candidates or those advocating extreme positions appreciated by a fairly small minority of the membership to win AB and TAG elections. We might well benefit from the diversity of such perspectives in the short run, but ensuring that other perspectives are also cycled in and out would prevent paralysis if partisans of competing points of view are elected.

[tantek]

@dwsinger appreciate the random thoughts, here's attempts for a few:

is there such a strong incumbent-advantage that better-qualified people are standing but failing to get elected (the problem that term limits purport to solve in public elections)

Close but not quite the actual problem we are (anecdotally) hearing.

Actual problem: there such a strong incumbent-advantage that equivalently-qualified people are not bothering to stand because they 1. (fairly accurately) perceive that incumbent-advantage, and 2. do not wish to suffer the humiliation (peers, professionally, by their employer) of standing and "losing". Whether or not they should feel that (they shouldn't obv) is not the point. The point is it is a human reality, and thus should be considered in our "software for humans" we call the process. Talk to Chris Wilson @cwilso or Andrew Betts if you want a more personal understanding.

surveys I have seen also suggest that people want term limits and want to re-elect their own representative

You're missing the (likely in several existing examples) people (orgs) that have others they would likely run if/when a "representative" termed out.

why we don't require that appointees in year N stand for election in year N+1...the Director could still appoint them if they fail to get elected

yes to your line of thinking here. makes complete sense, please file a separate issue to track that (perhaps something like "Should TAG appointees be automatically nominated for the next TAG election?")

[dwsinger]

see #166 ...

[tantek]

@michaelchampion makes additional good points, that I'd like to expand on:

There's a dilemma -- the most QUALIFIED people on both the AB and TAG

Yes to this entire paragraph, AND a system with term limits would strongly incentivize those with more experience (and desire to keep applying it) to mentor and knowledge transfer to newer members, and that can only be a good thing (broadens / strengthens institutional memory).

Not to beat my favorite dead horse or anything ;-)

Pretty much agreed with all that too. And it's not dead, not as long as we have contradiction in the process, and the possibility of electing "No other candidate" (hence the reason for keeping those two in the process).

[dwsinger]

This seems like a policy issue to start with. Referring to the AB (and they should probably call in the AC).

[tantek]

I’m happy to continue the discussion of @w3c AB term limits in the AB itself, however I’d prefer to keep this issue open to gather public feedback, comments, support, concerns, etc.

(Originally published at: http://tantek.com/2018/052/t1/ab-term-limits-issue)

[dwsinger]

yes, leaving open here as a possibility for future consideration.

[dwsinger]

we might also consider whether term limits for elected and appointed should be the same, or could be different.

[dwsinger]

AB's current thinking is maybe 3-4 terms, but we want to understand what we're trying to optimize for (the 'why?')

[frivoal]

Actual problem: there such a strong incumbent-advantage that equivalently-qualified people are not bothering to stand

I absolutely agree that this is the case.

To some degree, variations around the voting system may help, but since the incumbent advantage is also grounded in the incumbent's experience (i.e. it is not only an artifact of how people vote), it won't go away completely.

But a lifetime is a long time, and I would not want to block someone who's been a valuable member of the AB or TAG to be prevented from ever coming back to service if the AC felt they were the best for the situation at some point in the future. Also, once other people have had the chance to get elected, the incumbent advantage has faded out, so allowing the candidates back in seems alright.

This makes me lean towards limits to consecutive terms, rather than a lifetime limit. Maybe maximum 6 years in a row, or maximum 6 years in a 10 year period?

[LJWatson]

It seems to me, that we want an AB (or a TAG) with a balance of people with historical knowledge and experience, and people with new perspectives and experiences. If we exclude either, we're likely to be missing something valuable.

Anecdotal feedback is useful, but it also helps to have some actual data, so I did some archeology and pulled together the following information about the AB over the past 20 years.

41 people have served on the AB in the last 20 years (counting from the 1998 election to the present):

• 19 people (46%) served 1 term or less.
• 8 people (19.5%) served 2 terms
• 4 people (10%) served 3 terms or less.
• 4 people (10%) served 4 terms.
• 3 people (7%) served 5 terms.
• 1 person (2.5%) served 6 terms.
• 1 person (2.5%) served 7 terms.
• 1 person (2.5%) served 8 terms.

This means that 75% of people served 3 terms or less, or to put it another way, the majority naturally met the suggested term limit.

The current AB is a remarkably similar reflection of the historical data:

• 2 people have served 1 term or less.
• 3 people have served 2 terms
• 2 people have served 3 terms or less.
• 2 people have served 5 terms

In other words, 78% of the current AB have served for 3 terms or less.

[michaelchampion]

“Anecdotal feedback is useful, but it also helps to have some actual data,”

Thanks for the data, They don't address @tantek and @frivoal’s point that eminently qualified people are not bothering to run because incumbents who run usually get elected.

From a quick scan of AB election results, incumbents who run usually win. The years such as 2013 and 2014 where incumbents lost elections are also the years where there were the most candidates.

@frivoal notes that term limits should not imply a lifetime ban on running again. Right, the proposals I’ve seen just ask long term incumbents to sit out an election or a term.

[LJWatson]

@michaelchampion:
<From a quick scan of AB election results, incumbents who run usually win. The years such as 2013 and 2014 where incumbents lost elections are also the years where there were the most candidates.<

Which makes me wonder if a better approach is to encourage more people to stand for election (as we're discussing in #31). It would help us address the encumbent problem, help us maintain (or even exceed) the level of natural churn we have already, but without the risk of excluding people with useful experience based on a (more or less) arbitrary term limit.

In my single term on the AB, there have been numerous times when the two longest serving members have provided useful insight based on historical perspective, or better still, have saved the AB time and energy by summarising similar discussions that took place in the past. Without them, I think the AB would lose an important counter-balance to the other seven (all of whom have served three terms or less).

[fantasai]

+1 to Léonie's comments. I would argue that hard term limits run the risk of decreasing diversity of experience level, which is also important to have in a well-functioning committee--having me and Tab and dbaron each time out of the CSSWG after 6 years would have severely hampered the CSSWG, for example. We are especially valuable because we have been around so long. Advantages of such experience include not just history, but also continuity and mastery.

Some alternatives to consider might be a quota-based term limit system (where there's a limit on how many candidates past the term limit can qualify, and the election scores are used to determine which “very experienced” candidates are disqualified) or an enforced 1-term sabbatical after X consecutive terms, where X is high enough not to disrupt the group even when it's applied to particularly critical members.

But it might also be best to investigate less inflexible solutions than limits first.

[dwsinger]

Closing as we are not reaching consensus, and term limits have significant drawbacks.

[frivoal]

@tantek I am compiling a DoC for this year's process iteration. We resolved (at the Tokyo AB meeting) against your suggestion to introduce term limits in Process 2019. I'd appreciate if you could confirm you can live with that resolution.