[WikiEN-l] Is editing for payment a fundamentally problematic conflict of interest?

[William Pietri]

Delirium wrote:
> William Pietri wrote:
>   
>> Further, in many ways, we are in the same situation as journalists. Our 
>> product only has value to the extend that we have the trust of the 
>> public. That means that we need both to actually be trustworthy and to 
>> avoid situations whose appearance would undermine trust. Journalists 
>> have been wrestling with these issues for many decades, and we can and 
>> should learn from them. That's why I posted the relevant snipped from 
>> the SPJ ethics code.
>>   
>>     
>
> I don't think our situation is particular similar to journalists'.  We 
> are not doing original research, and we are not writing bylined 
> articles.  We summarize sources neutrally, post it publicly, and other 
> people edit our work mercilessly at the slightest hint of a problem with it.
>   

I think there are two questions here. One is about producing good work, 
and the other is about avoiding suspicious circumstances that would 
reduce public trust. I think you've addressed only the first, so let's 
start with that.


I agree that we're not doing original research, but we are still doing 
research. As you can see in [[WP:COI]], we ask people with conflicts of 
interest to not edit, for basically the same reasons journalists don't 
take payment or gifts from people they cover. (By the way, I'm not 
seeing how a byline works into it; even, and perhaps especially, for 
publications that aren't bylined, the ethics issues are still the same.)

When you describe the theory of how we work, I don't have a big argument 
with it. Our transparency and openness are what make Wikipedia possible. 
But in practice, I think that payment changes a balance that we depend 
on: that enough people with opinions on all sides of an issue have 
roughly equal amounts of time and energy to put into an article.

In areas where that's not the case, I think we can end up with pretty 
bad articles. Take a look at [[PA Consulting Group]], for example. 
Written in large part by PA staff and sourced mainly from press 
releases, I see it as a puff piece. But they had a lot more time to put 
into it than anybody else, and so they managed to [[WP:OWN]] it. Or to 
take another example from my watchlist, look at the fringe religious 
group [[Eckanar]]. I'm told by user Andries that it has always been an 
SPA playground. Just yesterday I hit my 3RR limit removing suspicious 
SPA edits, and nobody else lept into the gap. Will anybody have time to 
clean it up? We'll see, but given that the article was started in 2003, 
I'm not expecting anything to happen soon.

Or just look more generally at the topics where spam is a problem. The 
number of people with an interest in promoting their pet company or band 
or theory are much larger than the number of people around to work on 
articles like that, so we jump on anybody like that with both feet. 
That's not because we think them particularly bad people individually, 
but it's the only way we've found to cope with the vast imbalance.

So as far as doing good work goes, I think that avoiding conflicts of 
interest is just as important for us as it is for journalists. And I 
think, per[[WP:CSB]], we'd be wise not to introduce another source of 
systemic bias.


Even if the quality of the work weren't in question, though, we'd still 
come to the question of appearances.

Take a look at the reception reports get from pretty much any think tank 
or self-labeled institute. The reaction depends a lot on who is funding 
them. Global warming research funded by Exxon is not treated the same 
way as independently-funded research. Or look at the controversies about 
drug-company-funded research. Academics who consistently take a lot of 
money from a particular source are treated with a level of suspicion 
well beyond the norm.  Note that this doesn't even imply that any given 
academic is corrupt. Companies are more likely to fund people who are 
likely to make them look good, so there's a concern about selection bias 
as well.

All of this applies, by the way, to the academic realm, where people are 
putting years of training and their professional reputations on the 
line. The incentives are stronger for them to get things right. The 
incentives for J Random Contributor are not nearly as strong. 
Wikipedia's reputation is not nearly as well established as, say, 
Harvard. And neither is our internal governance.

So I think even if every one of the edits from paid contributors were 
very good, I think people would inevitably trust Wikipedia less. And 
given that I don't feel like we have a huge surplus of public trust yet, 
I think we should error the side of caution here.

> More similar, I think, would be to compare historians who write works on 
> commission.  These are generally paid for by an interested party, but 
> with the money given up front with the understanding that they're 
> commissioning an independent historical analysis that will not 
> necessarily show them in a positive light.  

I'd be intrigued to read more about this, but my guess is that it would 
require several conditions for it to work:

   1. The company would have to have a clear and special interest in
      being seeing as completely forthright.
   2. The historian would have to be somebody with an established
      reputation and solid credentials.
   3. The historian would do a relatively small amount of work for the
      commissioning party. (E.g., they would not be a staff historian.)
   4. The historian would not primarily do commissioned work.


Note that 1 is not the usual order of things. Companies spend tremendous 
amounts of money on PR and advertising, and vanishingly small amounts on 
historians. I'm sure that after WWII German companies had a big need to 
come clean, but that's not how business normally works.

Items 2, 3, and 4 are about reducing the conflict of interest to 
manageable proportions. Serious historians have a reputation to 
maintain. This is a situation notably lacking for freelance commercial 
writers. By allowing unfettered contributions from people with obvious 
COI problems, we would be much closer to the latter than the former.

William


-- 
William Pietri <william at scissor.com>
http://en.wikipedia.org/wiki/User:William_Pietri