[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Taler] Money with capabilities

From: belen barros pena
Subject: Re: [Taler] Money with capabilities
Date: Fri, 20 Aug 2021 09:41:01 +0100

Hi Özgür,

Thank you so much for posting about this. It is a fascinating subject, but it is an extremely hairy one too.

I feel I should give some context to my answer, which is somehow critical of the approach outlined in the document as I understood it. Over the past years, I have been doing research on the unintended consequences of digitising personal finances. I partnered with people who have experience of both mental illness and financial difficulty. This partnership uncovered that financial technologies somehow end up

1) prioritising efficiency above all other considerations;
2) removing all spending friction;
3) constraining the flexibility of the financial products and services we have come to rely upon;
4) indiscriminately collecting enormous amounts of very sensitive data about our lives;
5) creating extra work for us when managing our money; and
6) shifting power and control from us to the institutions that control the technology.

These are very worrying outcomes. I believe that, in everything we do as part of Taler, we should be alert and on the watch for these unintended consequences, and that we should strive to minimise them and counteract them as much as possible.

I also think it would help if financial technologies attempted to promote an alternative set of goals:  flexibility, complementarity, reflection, appropriation, positive security, collaboration and participation. In addition, it may help if we attempt to design these financial technologies from a social justice perspective.

My critique is mostly about 2 things:

1. The removal of UPP and
2. The power dynamics embedded in the design as described.

I will explain each in turn, proposing an alternative approach at the end.

1. The removal of UPP

If we look at fiat currency from a social justice perspective, what you refer to as UPP becomes a great equaliser. No matter who you are, if you have money in your hand, you can do and get all the things money can buy. In doing so, money treats us all equally. This is a remarkable property, and one that, in my opinion, is worth preserving.

I need to take issue with some of the arguments against UPP. 1.3 , the "reduced welfare" argument, I find particularly pernicious. In accepting it as a valid argument, we lend credit to the prejudices and the fallacies the argument is built upon: that people who have no money and require support from the state are reckless and incapable of managing their own lives; and that they are solely responsible for the circumstances they find themselves in. This is simply not true. Research has demonstrated numerous times the extraordinary financial acumen of those who have little money, and the fact that financial difficulty is most often derived from life events and systemic issues (e.g. low-paid, unstable employment), rather than personal behaviour or abilities. I think we should not be entertaining this argument at all.

Regarding 1.4, I am not sure we can say it is UPP what is "assisting personal harm and financial ruin". For instance, people with cognitive impairments can often handle physical, UPP money in day-to-day transactions. It is the technologies that currently mediate access to money what cause trouble: ATMs, phone and digital banking, bank cards and their attached PoS devices, complex products terms and conditions, and so on. They cause trouble not because of the cognitive impairment, but because they are poorly designed. In the case of people battling addiction, once again the role of UPP is probably negligible. Addiction is an extremely complex situation where numerous factors are at play. The truth is that eliminating UPP from money is unlikely to have any effect whatsoever on addiction circumstances.

I would therefore argue that we should aim for a design that offers the new capabilities you suggest, but that preserves UPP at the same time. Perhaps this could be done by conceiving your proposed features no as a set of special moneys with reduced capabilities, but as a new form of money: configurable money. The options you describe should not be restricted to specially-minted coins: they should come with all minted coins out of the box. We can then, as individuals, choose whether to activate those special configurations or not after the coin has been minted. This approach would avoid the development of special moneys: all money would remain equal at issue. It is individuals who, at their discretion, choose to configure their money in certain ways.

You may argue that this does not make much of a difference in practical terms, but I think it may. This is because special moneys are subject to prejudice and stigma, and may be rejected because of that. When I read how

"a minting authority could give out money with capabilities "age 14+" and "alcohol -5%""

I immediately wonder about the questions people may ask themselves when confronted with their special moneys. What if my 17-year-old friends find out my dad gives me kiddies' money? What if my employer found out that I use money issued by a "non-alcohol" exchange? What does that say about me? What kind of assumptions will the use of that money feed? In the same way as certain technologies created for older adults are rejected by their intended users because they are for "old, fragile people", we may come to reject many of these special moneys because they are loaded with negative social connotations.

2. Power dynamics

I think it is very important that, when we design a technology feature, we consider questions of power, and I worry about the power dynamics embedded in the design as described. This excerpt here illustrates why:

"Parents should be the authority that chooses the appropriate age group for the pocket money they are giving their children.  Similarly, legal guardians should be able to make such decisions for their wards."

This is, from my perspective, a quite troublesome approach. The reason being that it places all power and control on guardians. In doing so:

1) It enables abuse, both from institutions (e.g. refugees being given limited moneys along the lines of https://www.privacyinternational.org/explainer/4425/what-aspen-card-and-why-does-it-need-reform), and from individuals (e.g. abusive partners limiting the economic freedom of their spouses through limited moneys),

2) It reproduces many of the existing problems in legal forms of financial guardianship. Designed in this way, our "money with capabilities" may be used in coercive ways, and to enforce rules against the ward's wishes. One may argue that parenthood requires such things occasionally, but parenthood is rather extraordinary as guardianship situations go. First, children do not have all the rights adults have; and second, it is reasonable to assume that parents have their children's best interests at heart (at the very least, it is so in the vast majority of cases). Things get murkier when guardianship involves adults. These can find themselves forced into accepting financial guardianship arrangements against their will. For instance, in the case of representative payeeship in the US, people may lose their welfare benefits if they refuse. In the case of older adults and power of attorney, objecting to the arrangement may compromise the older person's only source of care and support.

We should not build technologies that easily enable abuse. At the very least, we should make abuse hard, if not completely impossible. Even if authoritarian governments do not need Taler to enforce financial censorship, do we really want to contribute another weapon to their arsenal?

Additional issues in relation to power dynamics surface towards the end of the document, when describing the safeguards built into Taler to avoid being locked out from legal access to UPP money:

"Peer-to-peer transactions could be used to exchange coins with limited capabilities for coins with UPP for a fee"

This arrangement exacerbates the disempowerment and vulnerability of those who find themselves in possession of limited moneys. The weakest party here (e.g. the refugee who has received only food money and needs to get some cleaning supplies or clothes) is made even weaker by enabling profiteering and predatory behaviour from those willing to "sell them" full-featured money. If a technology will make the weaker party to a power relationship even weaker, it should not be built.

"Based on trust, another person with access to UPP-money could simply share their (UPP) coins in exchange for the coins with limited capabilities."

I find this not to be a feature, but a workaround. Those who have no other option will be reduced to beg for favours in order to get full-featured money. Once again, we see a further weakening of the weaker party to a power relationship.

3. A possible alternative

Does this mean we forget the idea of configurable money? No at all. We perhaps take a more cautious approach, grounded on 2 principles:

1. No special moneys. As mentioned in section 1, these new capabilities are not added to certain coins at minting time, but they become part of the minting process. All money is issued with configuration capabilities: it comes like this out of the box. These configuration options can be enabled and disabled at the discretion of the coin owner after the coin has been minted.

2. Coin configuration must happen at the edge. The person who will be spending the coins must be the one configuring them. Transferring a configured coin to another wallet removes all special configurations. This would eliminate a whole chunk of possibilities for institutional abuse, and would make abuse by individuals somehow harder. My abusive partner would need to bully me into configuring the coins or, most likely, take hold of my device. We know vulnerabilities that require physical access are more difficult to deploy: I guess the same principle applies. Of course, this principle could also enable workarounds to invalidate the coin configuration (transferring the coins to another wallet that I also control), but at least this workaround tends to enable the weaker party.

The principles I have outlined seem to get in the way of the parental controls use case, but only in appearance: they just change the power dynamics and the behaviours involved in parental control. Rather than absolute power-yielding by parents, what the design encourages instead is collaboration between those involved in the coin transfer. As a parent, I can no longer send my 16-year old a coin with limited capabilities so that they cannot buy hard alcohol. I will have to sit down with my children instead, and convince them to configure their coins in ways appropriate to their age. That gives parents the chance to explain their reasons for constraining their children's behaviour; and gives children the chance to object.

Once the coin is configured, perhaps the configuration is final and cannot be changed (unless transferred to someone else's wallet of course). Perhaps changing the configuration is allowed but comes with a delay (e.g. 24 hours), so as to give someone time to think (do I really want to betray my parents' trust? Do I really want to place that online gaming bet? Do I really want to buy that thing?); or with a notification to the wallet that transferred the coin so that dad knows I've changed something in the money he sent me (but perhaps not exactly what the change was, so that dad will have to ask).

The same principles apply to relationships of financial guardianship between adults, encouraging a more collaborative approach. The process of configuring the coins may become an opportunity to discuss the wards' financial practices and habits, and a chance to engage the ward in the management of their own money.

Outside guardianship situations, configurable coins could become a tool for those who feel the need to exercise restrain and self-control when it comes to money (i.e. most of us). One of the more powerful examples I encountered during my research was from a man in his late 50s who lived with bipolar disorder. He was perfectly aware of the fact that during manic phases he overspent on travelling (trains, flights and hotels), and was experimenting with different ways of creating spending restrictions to handle the problem. One of the things he was trying was Monzo's lockable saving pots (https://monzo.com/blog/2018/12/13/locked-pots), so that money was not so easily accessible during these manic phases. Configurable digital money could play a similar role, contributing to people's personal money strategies.

There is a third principle I would propose: and it is that we do not simply roll out this kind of features, but that we strive to do research on them to fully understand their implications, so that we can roll them back if needed and explain why we did so. Hopefully this doesn't sound too outlandish.

Finally, you may argue that implementing a feature like the one I've described is too difficult or impossible. But please keep in mind that what I've described here is not really intended for implementation. It just wants to present a different perspective, and perhaps make us think about our features somehow differently. These are wicked design problems we are dealing with, and we should proceed with caution :)



On Wed, 18 Aug 2021 at 15:07, Özgür Kesim <oec-taler@kesim.org> wrote:

reply via email to

[Prev in Thread] Current Thread [Next in Thread]