**Introduction and definitions**
In the article to be published in the GoodIT conference 2022, we discuss the results of the analysis of transactions of the Circles UBI system. The analysis conducted helped us to respond to the following research questions:

* *How is the tokenized basic income spent by Circles users in the Berlin pilot?*
* *Who are the most prominent users of the Circles UBI system?*
* *How did the subsidy program affect the economic network structure in the context of the Circles UBI pilot?*

The time-span considered for the analysis is:

* **from October 16, 2020 to March 21, 2023.**

Our analyses targeted the overall network of the Berlin pilot. The Berlin Pilot network is the largest subgraph in which all users are connected with one another (directly or indirectly) containing the subsidized businesses.

In order to gain deeper insights on the subsidy program impact, we analysed the pilot in 3 emergent levels:

1. **The subsidized business network (B2B)**: this network includes *subsidized businesses* only (18 in the period considered), and the transactions taking place among them (714). This is the Business-to-Business network within the subsidy program. Businesses that do not get subsidized do not appear here. inesses that do not get subsidized do not appear here.
2. **The subsidy program network**: this includes both *subsidized businesses* and their trade partners (*UBI participants (subsidy)*), with a total of 675 users. These are defined as the group of users having directly transacted with subsidized businesses at least once during the time-span considered. Within this sub-network, a total of 5,993 transactions were performed in the period considered.
3. **Berlin Pilot network**: the total users of the Berlin pilot were 5,151 and the total number of transactions during the whole period (October 2020 to March 2023) were 31,035 transactions. This network includes all types of users, including the *subsidized businesses*, users part of the subsidy program *UBI participants (subsidy)*, and those who aren't connected to the subsidy program directly nor indirectly (UBI participants (P2P)).

We summarize our findings below.

**Circles UBI spending**
* We defined 7 categories related to the type of activity performed by the business:
* Food and Beverage
* Art
* Bikes
* Care/Health
* Co-Working
* Books
* Consulting.

Furthermore, by assigning the "cash-out" category to the Circles Coop wallet address, we calculated the amount cashed out by each of those business categories.

**Figure 1.** Transaction volume involving the subsidized businesses classified per business category (on the left).

**Results**
* Both the highest volume (389,101.29) and the highest number of transactions (3,183) involved the Food and Beverage sector.
* The Circles Coop transacted the second largest volume (349,972) with 876 transactions.
* The third spending category per volume concerns Art, towards which 678 transactions were made counting a total spending of 111,029.55 Circles.
* Other businesses involving significantly lower incoming trade in Circles tokens include:
* bike rental shop 60,943 Circles tokens , 297 transactions,
* businesses related toCare/Health 54,549.59 tokens, 291 transactions,
* Co-working 25,666.44 tokens, 147 transactions,
* book sales 17,363.11, 108 transactions
* Consulting 15,280.09 tokens, 24 transactions.

A focus on how the baisc income was spent by the businesses reveals that the vast majority of the transactions performed by the businesses was directed to the Circles Coop, which enabled the cash-out operation.

Only businesses belonging to three sectors reinjected a significant portion of their income in the system.
* Food and Beverage (37427.05 to other businesses and 9922.82 to UBI participants (subsidy))
* Care/Health (2396.15 to other businesses and 1702.53 to UBI participants (subsidy))
* Art(1742.37 to other businesses and 2640.44 to UBI participants (subsidy)).

**Prominent Users of Circles UBI**


**Figure 2.** Distribution of spending in Circles per user type.
Together UBI participants(subsidy) and subsidized businesses are the user categories among which the largest amount of volume circulates (they send roughly **62%** and receive **66%** of the volume).
The overall UBI in Circles spent to subsidized businesses (pie chart **on the left**) corresponds to **44.5%** of the total volume. The UBI spent towards UBI participants(P2P), to **34.3%**, and that spent to UBI participants(subsidy) amounts to **21.2%**.

The values change if we consider the source of the transaction (pie chart **on the right**): the UBI participants(subsidy) transact the largest volume (**44.3%**), P2P users trade **37.4%** and the subsidized businesses trade only **18.3%** of the volume.

**Analysis:** These results remark the importance of the program to facilitate the spending of users' personal currencies. They also show the subsidy program might have catalyzed imbalance of the payment system, because the subsidized businesses are mostly receivers and UBI participants(subsidy) are mostly senders, while UBI participants(P2P) play both roles rather equally.


**Figure 3.** Transaction volume per user type pairs.

| type | UBI participants(P2P) | Subsidized Businesses | UBI participants(subsidy) |
|-----------------------|---------------|-------------------------|------------------|
| UBI participants(P2P) | 752095 | 0 | 107825 |
| Subsidized Businesses | 0 | 346142 | 74077 |
| UBI participants(subsidy) | 37802.1 | 677363 | 304828 |

**Table 1.** Transaction volume per type pair martix.

Figure 3 and Table 1 show that the volume traded among UBI participants(P2P) is the highest.
Second is the volume traded by subsidized businesses to other subsidized businesses, and third, the volume traded from UBI participants(subsidy) to other UBI participants(subsidy). The remaining pairs have particulalry small traded volumes.

**Analysis:** These data highlight that the P2P user group formed trade communities independent from the subsidy program. These communities, however, were far less prominent in terms of trade volume compared to those involved in the subsidy program. The data further confirm the imbalance between trade from and to subsidized businesses.

**Structural changes in the economic network**

* The table below shows a remarkable decrease in the second period for the amount of users, number of transactions and trade partnerships in the overall Berlin pilot (from January 2022 to March 2023 on the right) compared to the first one (on the left between October 2020 to January 2022).

* For the Subsidy Program network and the Subsidized Business network, the number of transactions, amount of users and trade partnerships increased sharply from the first to the second period. This shows the growing influence of the subsidy program in the network dynamics of the pilot.

* The volume of transactions (the sum of all crc traded) increased in the second period across all 3 networks.
**Table 2.** General statistics on the aggregated graphs: each graph results from the split between the period before and after the Split-date shown. The resulting causal fidelity provides information on the accuracy of the static representations. The number of trade Partnerships result from the aggregation of the transactions in each period. The last rows indicate the number of weakly and strongly connected components respectively.
**Note** Trade Partnerships refers to the number of unique pairs of users who traded with each other in a specific direction. For example, A transacts with B once, and B transacts with A once. In this case, there would be two Trade Partnerships established.

**Analysis:** what this means is that the overall amount of users that join Circles irrespective of the influence of the Circles Coop has drastically decreased. The influence of the subsidy has meant that the people who use Circles are mostly connected to the subsidized businesses, although the volume of trade has increased for all users. The increase in volume of crc traded might be due to the fact that people have overall more crc to spend due to the monetary policy. Further research is needed on this aspect.

The monthly transaction volumes per user type and number of active users in the network confirm this hypothesis.

**Figure 4.** Number of active users. Users are defined as active in a given month if they perform at least one transaction in that month.

**Figure 5.** Monthly volume of out-going transactions classified by the user type of the sender.

**Figure 6.** Monthly volume of in-coming transactions classified by the user type of the receiver.

**Analysis:** As we can see from Figures 4 and 5, businesses are receiving more circles than what they are spending. We can also see that UBI participants(subsidy) tend to spend more than what they receive in crc. This means that there is an overall imbalance of payments from buyers, UBI participants(subsidy), to sellers, subsidized businesses. This is partly because of the fact that UBI is meant to be spent by overall users who are related to the subsidized businesses.

In Figure 7, a visualization of the network shows the structural change taking place in the network. Nodes are coloured by their type and ranked based on their centrality (the bigger in the visualization, the more central). In this context, the centrality of nodes is calculated based on the number of trade partnerships established. In the first observation, the most central nodes are those not connected to the subsidy program (blue nodes). In the second, by contrast, the most central nodes are either subsidized businesses (green nodes) or UBI participants(subsidy), coloured in red.

**Figure 7.** Berlin pilot network before (on the left) and after (on the right) January 2022.

**Analysis:** The network structure reveals that the distributions of in-coming and out-going trade partnerships change significantly from the first to the second period. We observe that the network transitions towards a more evident core-periphery structure in the second period. Moreover, in the second period, a larger portion of subsidized businesses and UBI participants(subsidy) become part of the main economic backbone, or core of the network, which is not the case in the first period.

**Centralised economic structure**

* In the first period there is less centralisation. In the second period, the overall network becomes more centralised in terms of the number of connections between different people & businesses. This means that few users have many more connections than the rest. However,the Subsidy Program network and the B2B network became more decentralized, showing the influence of the subsidy program.
* The amount of Circles flows controlled by the subsidized businesses increased exponentially in the second period, indicating their growing centrality in the network. However, the prominence of such users is not especially related to a role of intermediation or brokerage within the network.

**Analysis:** in the second period, **more people have more spending opportunities to use their circles** compared to the previous period among users connected to the subsidy program. In the rest of the network, the opposite is true. This shows the increased importance of users taking active part in the subsidy program.

**Notes:**
- We excluded from the analyses the intermediate steps of each transaction (if any), as well as all automated UBI distribution transactions to focus on user activity exclusively. The source address in our work is the wallet of the user who sent the transaction. Likewise, the target wallet is the final destination of the transaction. The amount of the transaction is, therefore, the amount transacted by the user pair.

Discussion with Cats and Product

- There has been a conscious decision from the cats to subsidize businesses to close loops, to sustain partners that offer products like food and drinks. e.g. consultancy -
- Define the difference between a non subsidy business partner and p2p network as we move forward.
- Understand better the p2p transfers and what is behind - e.g. UBI crowdfunding for therapy - try to understand the interactions that exist and how to foster them more

Sowelu:
- to improve the analysis and filter the noise related to the initial usage we can look at number of transactions and volumes after a certain threshold.
- locate the p2p network and ask questions for qualitative stuff

Alessandro
- history of subsidy program:
- Sof: at the beginning it was quite experimental - at the beginning the early birds were offered up to 1000 euro exchange per week (4000 per business per month), but it was a small amount of people so it was important to invite them to take a step with us. Then it was lowered to 3000 in Jnauary 2022 and then 2000 per month after July 2022 more or less. This was the moment where we onboarded a big chunk of the business network. Recently, when Sof joined, a formal way to phase out the subsidy was put into practice to allow for more businesses to join. How it works: for 6 months businesses receives a cashout of 3 categories - standard: exchange limit of 1000 , then to 800 after six months and then 500 - the idea is to work with them to think where they can spend this. 2nd category: 2000 euro per month, mostly for food partners and metamorfosis (health collective). 3rd category: reserve partners, who will have it hard to spend their circles e.g. tax advisors - has costs that cannot be covered with circles but all business partners rely on a tax advisor.
- In July we are starting the first decrease -
- Paying for labor in crc is currently being explored.
- Recently started with a business partner that does not have subisidy/does not exchange into euros.
- Carmela: people wanting to join the network without subsidy should be our concetration. Also, how to increase b2b transactions? risk of people only joining for subsidy - could be part of the questions for qualitative -
- people offering things is also the result of the transition - people are looking for ways to increase their UBI to be able to spend - this was probably a result of the transition from 240 crc/euro to 740 crc to 75 euros -. We are observing slowly more activity despite this fact so its important to focus on it for research.
- Currently in standby for new partners - for financial reasons but also for argumentative reasons: we want to do a more transparent and public outreach - "we want x and y businesses" -
- we want to identify which parts need subsidy
- Teo: for this we should have more information via the financial diaries to understand which complementarities we can find within the business network -
- Teodoro: we should create cycles according to business complementarity - we can try some kind of recomendation or matching algorithms to suggest/help cats to find businesses that are more complementary to each other in terms of production( what they sell, their inputs/outputs and the flow). We can try to find what is the best match to find cycles.

Lucia:
- categories and partners - constraint on the numbers - understand what is the way in which the UBI becomes useful /add on , depending on class, capitalist over-consumption? etc -
- explanation based on categories is more needed. for example: food - but what type of food? is roots radicals a first neccessity product? is this providing an extra consumption item - with coffee, people were buying 10kilos of coffee for free= is it needed? excessive consumption or basic income? who benefits from things being free?

Alessandro should reach to Carmela, who has a lot of experience with free-loaders.
carmela: when she talks to people, people say they cant spend all their circles - which also decreases their motivation - they cannot buy only roots radicals sauces - good direction on where to go with partners -
more research on What People need? of course people need food, which types of e.g. food?
about categories:
- art is misleading, needs to be remade - madame zorro does clothing and colorin does bowls and utensils.

pragmatic constraints:
- in monthly markets we need people who give products and have engagement and have an alignment of values -
- workflows and admin overhead due to technological development
- e.g. no english language means people who do not understand the language of the app cannot join and we do not have the time to do care for each individual user.
- timetable: important to have it to understand the "history" of what has been happening.

Q. to Alessandro:
- question of representativity - we should not only look at the mis-using ego centric use of circles - how do you want to choose who you want to talk to ? how to avoid biases?
- Lu: important for product: is not so binary - things are not counter-things? its not about what is a good/bad behavior but what is the politics behind? there is no research that is objective because there is always a position behind. What are the principles and the aims to pursue when doing the research.
- what you say basic income, what does it mean basic neccessities?
- analysis of fiat income to circles - is it at add-on? does it help people to finish the month?

Teo:
- guiding question: how can we help communities be self organised and work as indepently as possible for the official currency system? how we can create an economic community which as much as possible independent from the official system. Economic autonomy.
- ontological point of view: the intent of having an objective way of observing reality- true from psychological and socio-economic point of view. It is important to be as much objective as possible and 100% objective is impossible, we know, but nonetheless, we should find a way to agree what is green, red and blue or whatever. Its important to connect therefore qualitative and quantitative methods because when you have different ways of observing reality, we can check the limits of each method. Every method is incomplete, so therefore mixing them helps. We should not throw away methods because in the past they were used for political reasons - we should learn from every way of observing reeality and different glasses, specially when they bring pre-defined assumptions, because when they are ther,e we can test the assumptions and how to make it better, if its possible.
lu: is convinec the objectivity is a political lie - its not that she gives up on objectivity but chooses a different ways. They can be complementary -
- what are the pre-defined assumptions we base our research on? what are the sets of pre-defined assumptions we are playing with? based on pre-defined assumptions, its hard to have "objectivity".
How does teo's algorithm proposal would work: every user defines needs and offers - the idea is an algorithm gives u a list of possible matchings or groupings - it is a way of having an orientation of who is out of the system and maybe they have needs that can be satisfied by someone else but they dont know each other so probably these people who are out of the loop can be matched.
algorithm comments by Lu: the process how we developing this is tricky - then we are asked to fulfill the technology with content - product is trying to do it in a participatory way - how to make the algorithm participatory?

Sowelu - agent orient