THREAD: @paul_gertler & his co-authors conducted an experiment in which they threatened to shut off water to people in marginalized slums in Nairobi. Several commentators have raised serious concerns on Twitter about the ethics involved.
1/ https://twitter.com/paul_gertler/status/1292341505051381761

(I wouldn't have tagged Paul in this but @caiosborges has already tagged him into discussions about my concerns so he should at least know all of my concerns. I've little expectation Paul will also RT this thread.)

The article stems from economics but it relates to #humanrights & #bizhumanrights. It created adverse incentives that harm human rights of people in marginalized communities in order to seek better payment for utilities. The lead author works @WorldBank.
3/

I'm only going to comment on one issue re the article itself because others have addressed serious ethical issues elsewhere: lack of informed consent; using a fundamental human right as a bartering chip; adverse impact of the experiment on housing costs.
4/

You can read the concerns here:

5/ https://twitter.com/joshbudlender/status/1292170843389386761

And here:

6/ https://twitter.com/BernardsNick/status/1292365131939295240

I do want to highlight that the article brushes aside the fact that water & sanitation is a human right. You can read it in the pic. To be clear: there is no need for "" around right, just like one wouldn't say that "torture is a 'human right.'"

7/

While there are numerous ways one can define what is or is not a right, international law recognizes a human right to water and sanitation. Economists do not unilaterally get to dismiss that as a fact.

8/

https://www.un.org/waterforlifedecade/human_right_to_water.shtml

In response to concerns, they've posted 2 versions of an "explanation" to the ethical concerns. While the paper stems from economics, it implicates #humanrights & offers important lessons for #bizhumanrights.

https://twitter.com/paul_gertler/status/1292353595526979586

9/

So let's discuss ethics in experimentation in BHR using their explanation. The explanation was intended to solicit "feedback on how to improve the existing paper to better reflect the concerns raised."

10/

First, If an experiment is unethical, you cannot make the paper acceptable. Lack of informed consent irredeemably taints a paper.

The question needs to be whether the experiment was ethical or not.

11/

The authors note that they developed their experiment because a "culture of nonpayment became rampant" & threatened the ability of the utility to deliver water appropriately. The company needed *landlords* to pay their water bills to maintain quality service.

12/

Within #bizhumanrights, we need to know how to better provide human rights (no "") while addressing issues of cost distribution. I have no issue with such experiments. I do, however, have concerns about *starting* that with a human experiment that targets people/communities

in situations of vulnerability. One could do computer simulations. They could have used had Kenyan university students run simulations to determine what the likely outcomes would be.

They also, as @todipe suggested, could've engaged in community consultations.

14/

They chose instead to address this issue by testing whether threatening to turn off the utilities of *tenants* for *landlord's* non-payment would have an effect on revenue generation.

15/

I also should note now that I don't share @caiosborges's concerns about the academic pile-on on Twitter. The authors knowingly posted a version of a paper that they should have understood was controversial in order to solicit feedback. They got it.

16/

They might not like the feedback or feel it is unfair, but they got exactly what they were seeking by posting the paper.

If you're posting a #bizhumanrights paper that involves ethical considerations, you should be really clear about your ethics in the draft paper.

17/

They conducted a survey and determined that the issue was not that the tenants were refusing to pay but that well-off landlords weren't paying. Surveys are not consent to experiments.

Their solution to this problem was to target the tenants. Not the landlords.

18/

If you discover a problem is caused by the unwillingness of wealthy individuals to pay their fair share, your solution should not be to target already vulnerable individuals and make their life worse in the hopes that they will pressure the wealthy person in your place.

19/

I can conceive of no other explanation for developing this solution to target this problem than disdain for people in poverty or fear of upsetting those in wealth. If you hate poor people this much, or fear the wealthy, just stay away from these issues altogether. Please.

20/

I recognize I am ascribing bad motives here, and I don't like to do that generally. But this is the experiment: you discover person A is refusing to pay their share; you put pressure on person B, who is in a situation of vulnerability, in the hopes A changes their behaviour.

21

If person B is unsuccessful in changing person A's conduct, person B continues to suffer the consequences and seemingly nothing happens to person A.

There is no acceptable, good motive for this approach.

22/

The authors list several issues they monitored beyond repayment rates: access to water; child health; political activism; perceptions of quality & fairness. They do not list the impact on the physical or mental health of the adult tenants.

23/

Children are not the only ones entitled to the right to health. Anyone who understands poverty and has lived through it knows that receiving cut-off notifications dramatically increases stress, anxiety, depression, and other mental health issues.

24/

https://www.ohchr.org/Documents/Publications/Factsheet31.pdf

These impacts will often not manifest as political activism or a perception of quality/fairness. Poor people understand why cut-offs happen & will rightly blame their landlords. Political activism takes time, money, and often anger but not stress.

25/

Stress, anxiety & depression can be a deterrent to political activism. The authors don't seem to recognize that relationship or, really, understand poor people.

This is unsurprising. Let's look at the authors background:

26/

- No Africans authors, let alone anyone from Kenya
- All men
- All based in developed countries,
- All economists (no sociologists, anthropologists, human rights)
- Seemingly comfortable believing that a lack of community engagement/political activism is a good thing.

27/

If I were to put together a nightmare team for this kind of experiment, that would be it. This is not to demean the individual authors; it is not them as individuals that is the problem. It is them as a singular team that is the issue.

28/

If you are running an experiment in a developed country & you don't have a single co-author from that country, you need to re-think your team.

If you are running an experiment in Africa and there's not a single African (same for Asia & Latin America): rethink your team.

29/

If you are experimenting with poor people & no one on your team has ever been poor -- stay awake at night worrying how you'll pay your bills poor; crying because no matter how you calculate it, you cannot meet all your minimum payments poor -- rethink your team.

30/

This is a competency issue. If you are lacking that experience & perspective, you are lacking the necessary competency to conduct the experiment ethically.

31/

For the authors, despite their protestations in their explanations, this was an intellectual experiment & they did not design it with the requisite empathy for their human subjects because they did not have the requisite competency on the team.

32/

They experimented on 2 different human rights (water; mental health) with people in situations of vulnerability and without informed consent. I realize the IRBs likely approved this because it didn't alter the policy but simply structured who received cut-off notices.

33/

But it's still unethical to play with people's lives and human rights in this way without consultations or consent.

The authors try to pass the buck, noting that the interventions they used were designed by the government and are standard policies.

34/

This does not excuse the authors' choices. There is plenty of evidence that governments disdain the poor and fear the wealthy,. Following such a policy without question does not make a bad study better.
35/

Finally, I cannot take seriously any academic who writes this sentence:

"These interventions were not the invention of theminds ofacademics designed to test a theory, but rather to help agovernment sort out a real-world problem."

36/

In #bizhumanrights, most scholarship is focused on real-world problems. That does not change the ethical standards required of us.

An academic who writes this sentence has convinced themselves that their work & impact justifies any cost to their human subjects.

It does not.
//