Open Payment Standard

From EUOSR
Jump to navigation Jump to search

Introduction


As we enter a new decade of this still young millennium, sustainability has become an overarching goal for individuals and humankind alike. New ideas and approaches to ensure each person, corporation, and political unit can contribute to a more liveable world emerge every day. In seeking to make such a contribution to our collective future, we draw on one of the most established aspects of our economy and society: personal finance and consumption. As one of the main drivers of our well-being, both economic and quality of life, yet also a main contributor to the shared threat of climate change, the transformation of financial means into consumption is a key intervention point if we want to impact the future world in which we live. Doing so requires leveraging our technological progress in data science, and even artificial intelligence, so we can gain knowledge that changes our decisions and actions. Yet all this hinges on data and tools that are sensitive. Trust in the tools with which we bring about change will require them to capture the best of human machine interaction while offsetting our failings.


Our goal is to promote sustainable consumption by linking the so-called climate currency to actual currencies. People’s spending behaviour reflects their lifestyle. We seek to analyse individual consumption based on financial transactions, to provide carbon emissions for said transactions, and to thereby create transparency about one’s own individual carbon footprint. We then seek to enable financial institutions to turn this knowledge into actionable content, to clearly communicate information that helps empower people to change their consumption behavior and thereby reduce their carbon footprint.


Our approach is based on the idea of data transparency enabling consumers to become aware of their individual influence on global warming. It is our goal to give back power to these consumers, helping them to take control of their own environmental impact. With the individual's help we as service and goods providers can then alter our own behaviour, shaping our investment decisions with the proper foresight required to ensure they are sustainable not only in the short-term, but into the next millennia as well.

Table of Content

Our Purpose


The OfnK seeks to establish an ‘Open Standard for calculating consumer carbon usage on the basis of individual payment transactions’ (hereafter: Open Payment Standard). We are currently witnessing the proliferation of fintech firms specializing in many niche areas, as well as a growing interest by consumers in sustainable consumption due to concerns over global climate change. This co-occurrence has led to the first forays by financial firms/start-ups into using data to promote the participation of their customers in the fight against climate change.


While both these interests are welcome – fintech empowering consumers and consumers exercising control over our collective climate future – the technical possibilities raise risks if not implemented properly. Information is empowering, but it takes on the character of its user and what they choose to do with it: be it good or bad1. Thus data is sensitive: to cyber vulnerabilities, unreflexive mass collection, and shoddy data usage practices. Moreover, not all data is created equal, and standards for cleanliness, measurement, and accuracy can vary, with implications for outputs. While different algorithms can be used to deal with different data types and their weaknesses, algorithms raise their own red flags, the famous ‘black boxes’ making unexplainable decisions2. When considering this intransparency alongside the proprietary nature of most algorithms, the data input level seems the most viable intervention point to ensure the quality of technical systems. It is what we do with data that defines its value, and our actions, as good.


In any emergent field it is wise for the lead and early adopters (businesses, etc.) to engage in some form of self monitoring and regulation, before ‘normal accidents’ bring the attention of government regulators and ultimately customers, who impose external solutions: new legal restrictions or flight to another service respectively3.


Information is empowering, but it takes on the character of its user and what they choose to do with it.


The Open Payment Standard exists to promote such self-regulation. While it presents an Open Standard for the industry in these regards, it seeks not so much to set the standard, as to set the framework for discussions and collaborations across the financial sector and sustainability minded businesses. Only together can we develop an Open Standard in the interest of clients, customers, and climate. Within any broad strategic purpose, such as the above, it is important to set operable objectives, ones which can, within more-or-less existing ways and means, be realized and attained as stepping stones along the way to the grander goal. In this Open Standard we seek to establish the means and ways to attain what we identify as the core objectives needed to enable our strategic vision:

To reiterate, these objectives feed into the purpose of designing a widely adoptable, accurate, ethical, actionable, and agency-centric Open Standard for calculating consumer carbon usage on the basis of individual payment transactions.

Our Approach


Just because we can do something does not mean we should do something. This raises the question, why use individual standard payment transactions as a means to measure individual carbon footprints? Even if we accept preventing climate change as an inherent ‘good’, why do transaction-to-carbon calculating? How does this effectively contribute to the ultimate goal of a liveable planet?


The reasoning starts with the observation that individuals are the basis of community. Individual choices aggregate to determine the future of their community. Nonetheless, the community can and does exert significant pressures on the individual, directly or via nudging4. Yet even in the context of coercive or normative pressure, sustained and meaningful action by the individuals requires some level of voluntariness5. The purpose of a Transaction to-Carbon system is to promote such voluntary action. The hardest part of this task is already accomplished due to external context, that is, raising the individual’s interest in an issue. Climate change is a threat to survival and therefore receives increased attention6. Yet for this interest to turn into actions that contribute to survival, rather than to fatalism and apathy, another element is needed: information7.


This is where TtC systems come in: Information is the basis upon which a person can make decisions that lead to actions which meaningfully affect their environment, and feeding back new information that reflects how their actions changed the outcome drives continued and deepened action. This gives a person self-efficacy - a belief in, and evidence of, their ability to influence the world around them8. From the other side of the business relationship, this can be seen as a tailored customer empowerment experience that will lead to continuation and deepening of the business relationship with that customer as you maximize economic self-interest even in the pursuit of mutual goals (a liveable planet). This tension, once conveniently ignored in economic thinking that focussed on the profit imperative to the total neglect of social responsibility, is now at the heart of most business operations.


The resolution of this tension is best captured by the Triple Bottom Line (TBL) model9. It describes a balance between economic success, social responsibility (often referred to as Corporate Social Responsibility or CSR in the business world), and environmental consciousness. Through making visible the carbon footprint on an individual basis, the Open Standard and the associated technology makes it possible to put the theory of the Triple Bottom Line into practice: measuring CO2 emissions contributes to fighting climate change, while educating customers/citizens in order to make more environmentally conscious purchasing decisions contributes to societal development, community building, and the financial growth of sustainable businesses. The financial institutions adopting the Open Standard and the technology associated with it can in turn encourage their customers to invest in Environmental-Social Governance (ESG) funds and portfolios.


Information is the basis upon which a person can make decisions that lead to actions which meaningfully affect their environment, and feeding back new information that reflects how their actions changed the outcome drives continued and deepened action.

The Open Standard seeks to make a sustainable development agenda tangible in everyday action. Through consumer carbon calculations, it addresses the UN Sustainable Development Goals (UN SDGs) in the following ways: By firstly, making visible spending habits and, secondly, educating citizens about the climate impact of their spending behavior. The Open Standard and its associated technology address the specific Sustainable Development Goals and associated targets.


Figure 1. UN Sustainable Development Goals addressed by the Open Standard

SDG 8 Promote sustained, inclusive and sustainable economic growth, full and productive employment and decent work for all
SDG 12 Ensure sustainable consumption and production patterns
SDG 13 Take urgent action to combat climate change and its impacts
SDG 17 Strengthen the means of implementation for the global partnerships for sustainable development


Carbon footprint calculation is one of the most effective, quantifiable methods a) to understand a country’s, an organization’s, or an individual’s carbon footprint impact on the environment, and b) to involve everyone, from heads of states down to individual citizens, in actions to fight climate change. The Open Standard uses the most important indicator of sustainability - carbon footprint data - as the basis for optimizing financial institutions’ sustainable investment portfolios and as a means to promote behavioral change in individual consumers’ purchasing decisions.


Bank statements provide the data necessary for consumer carbon calculations and provide an important glimpse into individuals’ lifestyles, behaviours, and habits: frequent flights, online shopping sprees, luxury items, electronics, and so on. Even as sustainability and Corporate Social Responsibility gain importance as a pillar within corporate business strategies, individual consumers remain unaware of the extent to which they personally impact climate change - both positively and negatively. Despite their interest, they are not informed, even though the information exists. Financial institutions use this information to optimize their portfolios by, for example, granting social entrepreneurship loans, offering sustainable investment packages according to ESG criteria, and meeting the legal requirements regarding annual sustainability reporting10. However, maximizing the information to design a sustainability strategy and make informed investment decisions requires companies seek stakeholder engagement. Amongst these stakeholders, but often overlooked, are consumers.


Consumers possess a wealth of data alongside their power to move markets and drive change. Over time the analysis of consumer behaviour has become more fine-tuned. Yet there remains untapped potential. A business which treats consumers less as passive objects (to be marketed to), and instead recognizes them as the source of knowledge, power, and change, can leverage consumers as partners and radically improve its position in an improved world. This task - moving beyond considering persons as an abstract group of ‘consumers’ - and instead bringing together individuals, along with all the data and capabilities they possess, is a complex one. However, the complexity is fortunately able to be met with new tools developed for dealing with complexity: data science driven by information technology.


In the context of sustainability - and sustainable finance in particular - technology helps evaluate risks associated with social and environmental changes. For businesses, risk management has always been a vital part of their operations. However, the use of personalized information, rather than aggregate data, to tailor those investment decisions has not been mainstreamed in most financial industries. We see an exception in the insurance industry, where customer data, including data voluntarily provided by the customers, is used to cater offerings, adjust premiums, and guide corporate decisions. Such voluntary partnering between service providers and customers for the purpose of maximizing their interests, from quality-of-life to the liveability of the planet itself, is a fundamental principle of TtC systems.

Methods, Materials & Measurement


Transaction-to-carbon tools are made possible by modern data science. This generic statement has more complexity buried within it. Taking the broadest characterization possible, a TtC system will have three elements: data, algorithm(s), and hardware. Each of these elements has principles for best practice, but also some flexibility in their operationalization that requires decisions to be made and allows for a firm to specialize for comparative advantage.


The data is the element with the most aspects and decisions involved, being, as the popular contemporary metaphor goes, the ‘oil of the modern economy’. Several of the core aspects of data are closely linked to often overlooked decisions on hardware, such as how the data is collected and where it is stored and analyzed, both which also have security implications. Lastly, the algorithm is how data transforms into something meaningful for you and your customers, often called information or knowledge, which can be used to guide wise thinking and action, and then to inform future decisions, i.e. to gain foresight11.

Data Standards: Quality, Privacy, and Security


All developers and users of TtC systems will have their own ideas of what data needs to be collected to maximize the knowledge their algorithm(s) produce. These differences, necessary to distinguish a product and establish competitive advantage, complicate the development of data standards, but do not negate their necessity.


Standards are needed to ensure quality; allow comparison between products/offerings; facilitate compatibility, portability, and availability; and prevent peripheral or illegitimate purposes from being inserted into the TtC system. Most of these concepts fall into what is called information assurance, a wider conception of information management that includes but expands upon the typical security elements. These standards not only protect the system’s makers and users, but also assist in creating up and downstream markets that can magnify the value and impact of a product or service.


To ensure system quality, we need to first ensure data quality. Quality data is data that fits its intended purpose. The purpose of our data is to facilitate informed action. Thus a core principle is that information should be collected only if it has value in informing decisions and guiding actions. It is the character of these decisions and actions, not the data itself, that characterizes if data is ‘good’. That is, the data’s use, its purpose, defines it as good. The Open Standard ensures the use of ‘data for good’ by delineating data usage standards through common agreement amongst the ‘prosumers’ of that data for the purpose of contributing to the common good of climate protection.


Quality is often contrasted to quantity. Due to the low cost of data storage there is a tendency to collect and store all data possible given it may have future value. While big data has benefits, unreflexive mass data collection opens up two main risks. First is a legal risk, if the collected information exceeds what was necessary and/or explained in your data privacy policy. Second, large quantities of data increase your threat surface and thus security risks: more data makes you a more desirable target, increasing the probability you are subject to cyber intrusions. Being a lean target reduces risk.


Thus, only data directly necessary for measurement of the input variables necessary for your algorithm to produce information should be collected and stored, with exceptions for technical data (usage statistics, error reporting, etc.) or purposes clearly presented to your users in your privacy policy. In this regard we have one aspect which is less in tension than many others:


Security and privacy are linked. By focussing on lean-data collection, or ‘just-in-time’ data if you prefer, you can increase both.


However, this synergy is complicated by the fact that data retention is subject to certain legal requirements and customer demands. Data retention is often required for legally established durations. Customers/users want both to see their behaviour over time and typically also have the right to receive copies of their data (called ‘portability’). The complicated issues of data retention/storage sends one core message:


Limit the security and privacy risks by limiting your collection in the first instance to a needs-based principle, focusing on value of data along other dimensions (quality, depth, diversity, etc.) rather than pure quantity.


Privacy and security for a TtC system is greatly affected by the fact that the TtC systems will have several technical components - from servers to smartphones - linking together several actors: financial institutions, customer/consumers, firms, etc. This requires firstly much more thought about security of data in its three states: at rest (storage), in transit (between the smartphone, the internet, and the financial institution’s servers), and in use (analysis). Storage and analysis can happen in either the user’s device or the financial institution’s devices. There are benefits and drawbacks for all possible combinations of user-server/storage analysis options, ones which also need to factor in the risks and benefits of transit between locations. We will turn to these design choices later.

Algorithms


The algorithm(s) underlying a TtC system can do three things: analyze, learn, and extrapolate. Basic analysis based on data science and statistical methods is the bread-and-butter and does not entail what we today consider controversial design choices. It should, however, not be advertised to users as ‘AI’; it is not. The next level however would constitute a form of AI: using your algorithm to learn from the user’s data, both to improve the information provided to the user and to improve the information and foresight gleaned from the aggregate user data. Use of these types of algorithms should be disclosed as they can be intransparent in how they reach their conclusions.


The third category is also typically called learning, but is ideally separated out since it can be either statistical or machine learning (AI) based, and because it entails making assumptions in order to improve the analysis. These assumptions can improve data quantity, quality, depth, and even diversity, but also increase the uncertainty of the results, i.e. the probability that your assumption and thus conclusion is in error. To give an example: a more accurate distance for a journey can be ascertained based on the location of the next non-home zone purchase (what we call ‘Shanghai shopping spree extrapolation’).


The best way to deal with extrapolation is to make it user-driven. This has the added advantage of giving the user a task to increase efficacy, even if that is just hitting a button. Hereby they can translate money spent into some intermediate measurement variable (distance, location, volume), and then translate that into their individual carbon footprint. This allows consumers to see how their money transforms into carbon via consumption. It adds explainability to the algorithms and tangibility to the carbon emissions.

Methods and Data Approach


Calculating the carbon emissions of personal consumption is a new approach to assessing an individual’s climate impact. The number of such approaches is proliferating. Yet none of these approaches is correct in an absolute sense. As we do not want to contribute to the growing phenomenon of greenwashing, we have chosen to be completely transparent in our calculations, and if a scientifically valid value does not exist for a given variable, we do not use it. While in the medium-term arriving at correct calculations is our goal, achievable by using the collective data of the fintech community’s TtC systems, in the interim we seek to provide the best calculations possible by focussing on the relative position of a given consumer in the larger picture. While we cannot accurately state that person 1 has contributed X amount to global warming, we can get a good idea of their relative contribution compared to the average person or other persons.


In the following, we outline our four step approach for calculating a person’s transaction-to-carbon footprint, that is, gathering and structuring the data. The communication of this data as information will be turned to subsequently. The four steps are:


  1. Calculate the aggregate carbon footprint on the basis of publicly available data
  2. Set the right assumptions to break down the big picture into a more complete picture
  3. Personalize transactions with additional data (edge cases)
  4. Adjust the arrived at data to account for personal behavior.


These methodological steps are discussed in depth below.


6.1. Aggregate Data Based on Your Chosen Baseline


The information you provide to your customers provides them with knowledge about their relative contribution to the carbon footprint as compared to some form of baseline. As mentioned, providing some sort of absolute value is unfeasible bearing Orwellian data collection methods, if even then. The first step is to identify at which level of analysis that baseline should be. Baselines could be one of the distinguishing features of different TtC systems. However, best practice does require the baseline maximize variation even within an aggregate number. Carbon emissions are very dependent on the country / area in which people live. An avocado bought in Germany has a different footprint than one bought in Mexico (its country of origin).


We therefore approach carbon footprint calculations from a country level of analysis, allowing our aggregate outcome variable (carbon emission) to vary in a methodologically sound way for valid comparison and inference12. Specifically we developed this Open Standard focussed firstly on Germany, subsequently expanded it to other countries, and now with this document are opening it to community contributions from across the globe. Data for carbon emissions on a country level are easily accessible13. So are values for Gross Domestic Product (GDP). The first level of calculation is always based on these two variables.


Total Carbon Emissions / GDP = CO2 /€


This value is not very useful for gleaning insights into personal, individual spending for consumption, as it delivers values way below the actual personal impact. Thus, we need to next adjust this number for the individual level.


6.2. From Big Picture to Complete Picture


The next step uses statistical data about personal spending. Total national carbon emissions can then be divided by personal spending: Total Carbon Emissions per capita / Personal Spending per capita = CO2/€

This value gives a dedicated CO2 value per Euro spent. However, it does not take into account the different carbon values for different categories of spending, which needs to be done in another step.


6.3. From Complete Picture to Categories


The carbon footprint of different goods and services vary. A euro spent on fuel for your car has a different impact than a euro spent for a bus ticket. In order to include prevalent variations with regards to the carbon values, it is crucial to classify the goods and services into spending categories. The following list provides a possible classification into different spending categories:


  • Groceries
  • General
  • Service Stations
  • Energy
  • Local Public Transport
  • Train Travel
  • Lodging & Accomodation
  • Taxi Cabs & Limousines
  • Furniture
  • Clothes / Shopping
  • Streaming Services
  • Air Travel


Using research data from both independent and public institutions in the field of carbon emissions one can obtain values for these different spending categories. The combination of second level personal spending data per category and carbon values per category give dedicated values for the given categories.


Carbon emissions per category per capita p.a. / Personal spending per category per capita p.a. = CO2/€ per Category


At this stage we can already glean good insights into the different values in different categories. It does not, however, reflect individual consumer behavior such as diet, lifestyle, etc., but can form the basis for such calculations.


6.4. A Person Is Not a Category: Personalizing the Data


The goal of individualizing carbon footprints on the basis of personal spending patterns requires further elaboration on the different categories. Each category contains great variation within it. People’s idea of what constitutes a meal range from the meat platter to the vegan dish, and carbon impact varies along with these preferences. Thus, each base value CO2 unit needs to be adjusted by these personal factors:


Base value CO2/Unit * Factor 1 * Factor 2 * Factor X = CO2€/Transaction


These specifics, the heart of a good TtC system essential to turn the data into information relevant to a person, are outlined in the next section.


6.5. Categories


To begin personalizing, the categories themselves are subdivided into smaller categories, clusters, and merchants. ‘Accommodations’ is imprecise, a hotel versus a hostel may differ only by one letter but are far apart in environmental impact. These categories are presented in Figure 2. In further developing the standard, we will enter additional information as it becomes available in order to provide a better carbon footprint based on an even more fine tuned breakdown.


For some of the categories we are generally able to still use base values. Nonetheless, we are continuing to look for additional values for these categories in order to better estimate carbon emissions. For many of the categories though, the base values require personalized adjustments as discussed above and we have already developed calculation factors for these. These require extensive illustration, not only for their specific content but to illustrate to others how such personalization should be done for any variable in which they have interest.


Figure 2. Overview of Categories and subcategories

Category Subcategory Cluster (if applicable) Merchants
Groceries Catering
Bakeries & Cafés
Convenience Stores
General Online Shopping
Car Rental
Drug Stores
Travel Agencies
Service Stations
Energy
Local Public Transport
Train Travel
Lodging & Accomodation
Taxi Cabs & Limousines
Furniture
Clothes/Shopping
Streaming Services Streaming Services Video
Streaming Services Audio
Air Travel


6.6. Measuring Variables


Recapping: Identifying the variables needed to assess an individual’s carbon footprint based on their transactions is a critical first step, but meaningless if we cannot identify an accurate, reproducible, and comparable way to measure those variables as easily as possible. The transaction itself gives us one critical measure in an established consistent unit: currency. These measures can then be fairly simply attached to given categories of spending using existing methods. Specifically, matching a financial transaction to a category and thereby to a carbon emission is done in three ways:


  1. With Merchant Category Codes (MCCs) provided by the financial institution for credit card payments
  2. With service providers, such as FinTechSystems to categories SEPA / Direct Debit Payments
  3. By merchant name via the standard algorithm for debit payments / SEPA transactions if not categorized.


From here though, changing the transaction into a carbon outcome variable requires in most cases more elaborate measurement, including multiple input variables, some of which are what methodologists would call ‘fuzzy’ and may require subjective input from the consumer. Management of these variables should be guided by several standard principles: that the values are mutually exclusive and collectively exhaustive, and that the values have meaning in relation to the other values, that is they represent a scale from low to high even if they do not specify exact numbers or distance between any point on the scale. This means trying to avoid nominal variables, though we occasionally include this term when the precise order of a variable’s value is not straightforward.

Figure 3. Overview of Categories and Related MCCs and FTS Categories

Category Merchant Category Codes FinTechSystems Categories
Groceries/Retail MCC 1711, MCC 1771, MCC 5199, MCC 5411, MCC 5422, MCC 5441, MCC 5451, MCC 5921, MCC 5999, MCC 9751 K1.1.1; K1.1.2; K1.1.4, K1.1.6
Service Stations MCC 5172, MCC 5541, MCC 5542, MCC 5983, MCC 9752 K4.2.1, K4.2.2; K4.2.3
Local Public Transport MCC 4011, MCC 4111, MCC 4131 K4.2.2; K4.3
Train Travel (long distance) MCC 4112 K4.2.2; K4.3
Taxi Cabs & Limousines MCC 4121 K4.3; K5.3.1
Catering MCC 5811, MCC 5812, MCC 5813, MCC 5814 K7.1.1, K7.1.2; K7.1.4
Lodging & Accommodation MCC 3501, MCC 3502, MCC 3503, MCC 3504, MCC 3505, MCC 3506, MCC 3507, MCC 3508, MCC 3509, MCC 3510, MCC 3511, MCC 3512, MCC 3513, MCC 3514, MCC 3515, MCC 3516, MCC 3517, MCC 3518, MCC 3519, MCC 3520, MCC 3521, MCC 3522, MCC 3523, MCC 3524, MCC 3525, MCC 3526, MCC 3527, MCC 3528, MCC 3529, MCC 3530, MCC 3531, MCC 3532, MCC 3533, MCC 3534, MCC 3535, MCC 3536, MCC 3537, MCC 3538, MCC 3539, MCC 3540, MCC 3541, MCC 3542, MCC 3543, MCC 3544, MCC 3545, MCC 3546, MCC 3547, MCC 3548, MCC 3549, MCC 3550... further results K7.2
Air Travel MCC 3000, MCC 3001, MCC 3002, MCC 3003, MCC 3004, MCC 3005, MCC 3006, MCC 3007, MCC 3008, MCC 3009, MCC 3010, MCC 3011, MCC 3012, MCC 3013, MCC 3014, MCC 3015, MCC 3016, MCC 3017, MCC 3018, MCC 3019, MCC 3020, MCC 3021, MCC 3022, MCC 3023, MCC 3024, MCC 3025, MCC 3026, MCC 3027, MCC 3028, MCC 3029, MCC 3030, MCC 3031, MCC 3032, MCC 3033, MCC 3034, MCC 3035, MCC 3036, MCC 3037, MCC 3038, MCC 3039, MCC 3040, MCC 3041, MCC 3042, MCC 3043, MCC 3044, MCC 3045, MCC 3046, MCC 3047, MCC 3048, MCC 3049... further results K4.3; K7.2
Online Shopping MCC 5964, MCC 5969 K4.6.1; K1.1.6
Car Rental MCC 3351, MCC 3352, MCC 3353, MCC 3354, MCC 3355, MCC 3356, MCC 3357, MCC 3358, MCC 3359, MCC 3360, MCC 3361, MCC 3362, MCC 3363, MCC 3364, MCC 3365, MCC 3366, MCC 3367, MCC 3368, MCC 3369, MCC 3370, MCC 3371, MCC 3372, MCC 3373, MCC 3374, MCC 3375, MCC 3376, MCC 3377, MCC 3378, MCC 3379, MCC 3380, MCC 3381, MCC 3382, MCC 3383, MCC 3384, MCC 3385, MCC 3386, MCC 3387, MCC 3388, MCC 3389, MCC 3390, MCC 3391, MCC 3392, MCC 3393, MCC 3394, MCC 3395, MCC 3396, MCC 3397, MCC 3398, MCC 3399, MCC 3400... further results K4.2.2
Bakeries & Cafés MCC 5462 K1.1.4
Drug Stores MCC 5122, MCC 5912 K3.2
Convenience Stores MCC 5499 K1.1.4
Clothes / Shopping MCC 5137, MCC 5139, MCC 5611, MCC 5621, MCC 5631, MCC 5641, MCC 5651, MCC 5655, MCC 5661, MCC 5681, MCC 5691, MCC 5697, MCC 5698, MCC 5699, MCC 5948, MCC 5949, MCC 7245, MCC 7296 K1.1.5; K5.1.3
Travel Agencies MCC 4722 K5.4
Furniture MCC 5021, MCC 5712, MCC 5832, MCC 5932, MCC 5933, MCC 5937, MCC 7641, MCC 9753 K2.2; A4.4
Streaming Services MCC 4899, MCC 5815, MCC 5816, MCC 5817, MCC 5818 K5.1.2; K1.1.5
Energy MCC 4900 tba.


Our approach to further personalizing key categories, to turning them into variables that vary from person to person, is outlined in the following sections. Variables such as these are areas in which user engagement via their edge device (e.g. smartphone) adds essential value to the calculations while also enhancing the user’s efficacy. Examples of standard customer dialogues for obtaining this information are included in each category below. Dialogues can be profile or transaction specific, though profile specific dialogues will require intermittent review to capture changes in behaviour over time.


The dialogue illustrations also show how qualitative answers are transformed into usable data as standard variable types, specifically: continuous (includes both internal and ratio14) and categorical divided into ordinal, dichotomous (boolean), and nominal. Furthermore the chart also includes the selection method for each question-answer combination, important for data collection and user interface teams. The baseline number used in the example below is for Germany.


Figure 4. Adjustment Factors

Factor Unit Value/Weight
Duration in hours Flights up to 2 hours = short range

Flights of 3 - 4 hours = medium range Flights of 5+ hours = long range

133,170 gCO2/h

122,693 gCO2/h 132,503 gCO2/h

Class Economy Class

Business Class First Class

All ranges: 1

Short range: 1,3 Medium/long range: 1,9 Short range: 2,4 Medium/long range: 3

Route One way

Round trip

1

2


Figure 5. Calculation Formula

Base Calculation Sample Calculation for a 4 hour business medium range round trip flight
Base Value gCO2/h according to Range * Duration in Hours * Class Factor * Route Factor 122693 * 4 * 1,9 * 2 = 1,864.93 kgCO2

Understandable and feasible: from data to action


The first purpose of gathering data on individual consumer behaviour is to increase consumers’ awareness about their everyday consumption habits and how they affect climate change. As much as knowledge about the individual impact on global warming is valuable, it is only the first step in a long process of individual change in behavior towards lifestyles and habits that help improve or maintain the individual carbon footprint. This environmental awareness must transform into actions that improve people's lives while at the same time preserving our planet’s valuable resources. The provisioning of individual carbon footprint calculations, encourages banks and other financial institutions to conduct value-driven conversations with their clients and underscores the financial institutions’ commitment to sustainable development and sustainable investment practices - not just on paper but in practice.


Consumers dispose of strength in numbers. A single individual is seldom in a position to drive change globally, but when bank account owners jointly perform minor changes in their everyday behaviours and habits (i.e. changes in lifestyle choices like diet, means of transportation, etc.) the environmental impact can be considerable. Financial institutions can support these efforts by assuming the role of an ‘interface’ between merchants and consumers. The Open Standard hence acts as the connecting link or ‘glue’ tying these stakeholders together. When data is presented in an easy-to access and easy-to-use way, customers will discover added purpose and value in their everyday actions. However, this goal cannot be achieved with data alone. Communication is needed.


It is important for sustainability oriented financial institutions to engage experienced partners to assist in this task. These third-parties must possess expertise in the areas of technology, sustainability, and finance alike, in order to collect, calculate, and analyze TtC data. This partner must also be able to explain what the data means to financial institutions so as to enable them to engage with their customers in conversations on sustainable finance. Numbers derived from TtC calculations act as the basis for financial institutions to conduct such sustainability-oriented and value driven engagement with their customers. The TtC data makes consumer impact visible and offers the basis for providing consumers guidance on what concrete actions they can perform to help mitigate the effects of climate change.


This specific type of customer engagement in the area of finance can potentially develop into an added value of ‘Sustainability as a Service’ (e.g. in the shape of responsible investment decisions).


Bank account owners (i.e. consumers) will experience this additional service as a value add provided by their financial service provider. A partner or tool that empowers an individual with information and enables them to take actions, i.e. that enhances a person’s self-efficacy, is rewarded with the trust of that person, allowing the value added - to service, customer, and planet - to accrue even further over time15.

Law and Standards


When establishing an Open Standard that uses financial data as a measuring unit, certain laws, standards, and ethics must be followed. The Open Standard operates within a legal, ethical, and regulatory compliance framework that varies greatly by location. Moreover, the location of your corporation, the servers you use, and your customers, do not limit the extent of your legal compliance and liability16. For instance, while the EU’s data protection regime contains protections/limitations against transfer of a person’s data to non-EU states, every regulatory framework has exceptions of which service providers need to be aware. In this example, the EU-US Privacy Shield framework provides a mechanism for the transfer of personal data for commercial purposes between EU states and the US. Almost every country has extraterritorial legal tools as well, allowing law enforcement and/or private lawsuits across national borders. Diplomatic relations mean such cases will be given a fair hearing, even if their merits are low, and even if you ‘win’, the costs can be high.


Outsourcing some of your data management, a common practice, also will do little to ease compliance and liability responsibilities. For instance, even if you use cloud services, such as AWS or Google, these services are clear that you are the owner of the data; this makes you responsible for most routine law enforcement requests. These large companies will go to bat for you only in quite specific cases, usually when overly broad requests are made with precedent setting implications.


A universal good business practice, irrespective of borders, is – to borrow a term we used earlier – exercise foresight. It is important to not simply follow legal frameworks (a ‘Wait and See’ strategy) but to take a more active stance with regard to future developments - regarding law, environmental standards, and business ethics. This is the ‘Think Ahead’ approach. Take for example the existing ‘Do not track’ regulations applicable on the internet. Similar regulations will in all likelihood soon be transferred also to smartphone applications, and transaction data includes geolocation elements. Such legal transfer is not only inter-sectoral, but also geographic17. The most comprehensive data privacy regime is the EU’s General Data Protection Regulation (EU) 2016/679 (GDPR). This regulation, in whole or part, will influence new regulations in countries across the world. Thus even if a given firm does not operate within the EU, conforming with the GDPR today will prevent difficult and costly adaptation processes later: either when GDPR-style laws spread, or when a firm’s business operations expand as it grows.


Being at the forefront of future developments by implementing self imposed regulations offers a significant competitive advantage, also in terms of developing new business models and services. In this context, it is imperative for businesses to observe regulations in adjacent areas and thus deduce what influence they may have on your own business. Additionally, today’s standards are often the basis of future laws. For businesses, it is vital to adhere to international standards, so they are prepared for the legal framework of tomorrow. Of most relevance on the sustainability side is the United Nations Environment Programme Finance Initiative. UNEP FI works with more than 300 members – banks, insurers, investors, and over 100 supporting institutions – to help create a financial sector that serves people and the planet while delivering positive impacts.


In 2019, UNEP FI released ‘The Principles for Responsible Banking’, with the goal of providing transparency on how financial products and services create value for their customers, clients, investors, as well as society. They are part of the banking and finance sector’s contribution to achieving the UN Sustainable Development Goals (SDGs) and the Paris Climate Agreement. One more level up, the UN Global Compact is a non-binding pact between the United Nations and businesses worldwide. It seeks to promote responsible business practices by directly engaging with businesses themselves. Directly relating to sustainable finance, the UN Global Compact includes the ‘Action Platform Financial Innovation for the SDGs’, which seeks to assist financial institutions in ‘uploading’ UN policy into their financial services and portfolios by adding private finance to the global sustainability equation.


The Open Standard for consumer carbon calculations potentially closes an important gap between the aforementioned UN agreements and initiatives by putting one very important stakeholder at the centre of attention: educating and empowering the consumer who makes purchasing decisions on a daily basis.


As laws change, new international agreements emerge, or novel company-specific guidelines are formulated, the Open Standard will be modified to remain on the cutting edge, so that those following the Open Standard are ahead of the curve.

Options, Decisions and Guidelines


A standard should not be a straitjacket, but rather a flexible framework able to fit the needs of many different actors. In the above we outlined some important guidelines one should follow when making a TtC system, while we also highlighted areas in which innovators are able to distinguish themselves and make a ‘better’ TtC system. Options are necessary not only for innovation, but also due to differences in organizations. Specifically, any standard aiming for wide adoption must be customizable in how it can be implemented within an organization without sacrificing content. The most basic technological adoption matrix considers a technology along two dimensions: the financial intensity required for innovation/adoption, and organizational intensity required18. Within each quadrant a certain type of firm is most able to innovate, while also a certain direction in the technology is likely more beneficial. We illustrate the basics of this in Figure 6.


Figure 6. Technological Adoption Matrix

High Low
High Partnership space Big organization space
  • quantity & diversity of data
  • AI tolls (e.g. better chatbots)
  • server centered security
Low Small organization space
  • depth of data
  • responsive, rapid updates incorporating new knowledge
  • edge device centered (user control)
Individual (user) space


In order to avoid biasing the discussion of such principles before it has opened (the so-called anchoring effect) we ask directly proposing standards herein. Instead, we propose the following questions to guide the thinking of current and future designers and users of TtC systems and standards for these:


  • What impact will your system have? How will this impact be measured and communicated?
  • How does your system maximize the meaningful choices of customers? How do you plan to create feedback loops to ensure your customers can expand and deepen their engagement within the system?
  • How can you design purposeful sustainability data management practices? Exactly when and where do you need data?
  • To what extent will you use data to extrapolate, i.e. to reach conclusions based on both data but also assumptions? How will you control and guide these processes?
  • To what extent do you want to go beyond statistics and into machine learning? What transparency trade offs will be necessary?
  • How will you match data characteristics and machine learning types?
  • How will you monetize your system? Thinking about this in advance ensures the system is financially sustainable, your intentions are clear to your customers, and prevents mission creep later that changes the rules on your users as you desperately scramble for any way to turn a profit.
  • How will you monitor and incorporate ideas from different sectors, geographical areas, and non-binding frameworks to stay ahead of the legal and ethical curve?
  • How do we give zero meaning? The above outlined data inputs often assume an impact. Indeed, zero impact on the environment is a high benchmark. Yet increasingly new products, and new practices, are creating little closed-systems of near zero impact. Yet reporting these into a TtC tool is difficult, as this would require extensive customer dialogues.
  • What skills does your organization lack internally, but which are essential for your system’s success? Where can you get help? Who are suitable partners for supporting the implementation of a TtC tool?

Conclusion: Quo Vadis- where do we go from here?


The purpose of this Open Standard has been to start a conversation. We presented what in our view is a solid approach to building transaction-to carbon tools. Yet to make such a standard truly universal requires the engagement of more stakeholders: of NGOs, banks and financial institutions, fintech start-ups, and potential end users. It requires the input of coders, entrepreneurs, lawyers, psychologists, environmentalists, economists, and bureaucrats, amongst others.


Open participation by the whole community in developing this Open Standard leads naturally to one of the big outstanding questions: what other aspects of TtC systems should be ‘open’, that is, freely accessible to the community. Your data, algorithms, and programme code have value. However, the proprietary value may be exceeded by the value created if some of these are open source. Open source projects can be much more secure, accurate, and comprehensive, while also sending a message to certain actors - ‘hackers’ and consumers - about your intentions and ethics. Open source does not necessarily mean your end-product is free, which would be ‘Free and Open Source software’ (FOSS), though some innovators may wish to move in this direction and earn their profits through services19.


The second big question is how far to move TtC tools towards artificial intelligence. Traditional statistical methods work perfectly for many of the components of a TtC tool. However machine learning (the most popular form of AI at the moment) can enhance many elements, reducing the amount of busy work a user needs to do (e.g. less customer dialogues, or easier ones) and offer more tailored feedback, to the point even of being able to ‘discuss’ ideas for areas of improvement with the user (i.e. chatbots). Additionally, machine learning, as well as even some statistical methods, can allow a TtC system to extrapolate some conclusions from the data. The extent such extrapolation should be allowed in a TtC system is an open matter, though we have argued here that putting the user at the centre of extrapolation is best practice.


Technology is a tool for changing the world, but only by putting people at the centre of that tool can its use be truly transformative, changing the person so the changes in making the world more liveable are themselves sustainable.


Lastly, the community needs to further consider the user, their customers, people. While we have outlined how the output of a TtC system should be designed for effective communication, integrating people earlier in the ‘production of knowledge chain’ is a larger task. We do this presently by integrating their data. But people are not their data, their data is rather an incomplete picture of the person, “fragments” as put by tech pioneer and philosopher Jaron Lanier20. Let us take even just the broadest reason this is a problem. Change, in economic growth and environmental progress, is not driven by averages of the aggregates; it is uncertainty, exceptions, and outliers that drive the big things. These are difficult to account for in most formal systems of thinking and even harder to use as a guide for decisions and actions. Bringing user input closer into the production of knowledge, while not overwhelming them with requests, so that the advice and foresight offered is more significant and meaningful is a key challenge going forward. Technology is a tool for changing the world, but only by putting people at the centre of that tool can its use be truly transformative, changing the person so the changes in making the world more liveable are themselves sustainable.


References

1 Lehrke, J. (2017). Open Participatory Security: Unifying Technology, Citizens, and the State. Lanham, MD: Rowman & Littlefield, 4.

2 Pasquale, F. (2015). The Black Box Society: The Secret Algorithms That Control Money and Information. Cambridge, MA: Harvard University Press.

3 Malhotra, N., Monin, B., and Tomz, M. (2019). Does Private Regulation Preempt Public Regulation? American Political Science Review, 113(1), 19-37; OECD. (2015). Industry Self Regulation: Role and Use in Supporting Consumer Interests. OECD Digital Economy Papers, No. 247. Paris: OECD Publishing. http://dx.doi.org/10.1787/5js4k1fjqkwh en Perrow, C. (1984). Normal Accidents: Living with High-Risk Technologies. Princeton, NJ: Princeton University Press. Karpf, D. (2011). Open Source Political Community Development: A Five-Stage Adoption Process. Journal of Information Technology & Politics, 8(3), 323–45.

4 Thaler, R. H., and Sunstein, C. R. (2008). Nudge: Improving Decisions about Health, Wealth, and Happiness. New York: Penguin Books.

5 Stukas, A. A., snyder, M., and Clary, E. G. (1999). The Effects of “Mandatory Volunteerism” on Intentions to Volunteer. Psychological Science, 10(1), 59-64.

6 Selection bias towards threatening stimuli, see Bovaird, T., Loeffler, E., van Ryzin, G. G., and Parrado, S. (2014). User and Community Coproduction of Public Services: What Influences Citizens to Coproduce. In Public Administration and the Modern State: Assessing Trends and Impact, edited by J. Raadschelders, E. Bohne, and J. Graham with J. Lehrke, 109–24. Basingstoke, UK: Palgrave Macmillan, 123–24; Myers, J. R., Henderson King, D. H., and Henderson-King, E. I. (1997). Facing Technological Risks: The Importance of Individual Differences. Journal of Research in Personality, 31(1), 1–20; Panksepp, J. (2004). Affective Neuroscience: The Foundations of Human and Animal Emotions. Oxford and New York: Oxford University Press.

7 On withdrawal Girandola, F. (2000). Peur et persuasion: présentations des recherché (1953–1998) et d’une nouvelle lecture [Fear and persuasion: Review and re-analysis of the literature (1953–1998)]. Année Psychologique, 100(2), 333–76; Weinstein, N. D., Lyon, J. E., Rothman, A. J., and Cuite, C. L. (2000). Preoccupation and Affect as Predictors of Protective Action Following Natural Disaster. British Journal of Health Psychology, 5(4), 351–63.

8 Key work in this regard is Bandura, A. (1997). Self-Efficacy: The Exercise of Control. New York: Freeman. See also Breakwell, G. M. (2014). The Psychology of Risk, 2nd edition. Cambridge, UK: Cambridge University Press, 61; Lehrke, Open Participatory Security, especially 107, 195.

9 Elkington, J. (1994). Towards the Sustainable Corporation: Win-Win-Win Business Strategies for Sustainable Development. California Management Review, 36(2), 90–100.

10 In Germany, according to the CSR Directive Implementation Act ‘CSR-RUG’ issued in 2017, such reporting is mandatory for for-profit companies with more than 500 employees. See Rat für Nachhaltige Entwicklung. (2020). 11 “It is vital to remember that information - in the sense of raw data - is not knowledge, that knowledge is not wisdom, and that wisdom is not foresight. But information is the first essential step to all of these” (Arthur C. Clarke quoted in n.n. (2003). Humanity will survive information deluge — Sir Arthur C Clarke. OneWorld South Asia, 5 December. https://bazaarmodel.net/phorum/read.php?1,462).

12 King, G., Keohane, R. D., and Verba, S. (1994). Designing Social Inquiry: Scientific Inference in Qualitative Research. Princeton University: Press Princeton, NJ.

13 See United Nations. (n.d.). Greenhouse Gas Inventory Data - Comparison by Category. https://di.unfccc.int/comparison_by_category

14 The issue of ratio variables is complicated since zero is both hard to measure and hard to reach regarding carbon emissions. We will comment on this in the concluding section(s).

15 Heath, R. L. and Palenchar, M. (2000). Community Relations and Risk Communication: A Longitudinal Study of the Impact of Emergency Response Messages. Journal of Public Relations Research, 12(2), 131–61; Wojtczak, M. and Morner, M. (2015). Bringing the Citizen Back In: Motivational Aspects of Knowledge Sharing through Web 2.0 Technologies in Public Administration. eJournal of eDemocracy and Open Government, 7(2), 29–44. Peters, R. G., Covello, V. T., and McCallum, D. B. (1997). The Determinants of Trust and Credibility in Environmental Risk Communication: An Empirical Study. Risk Analysis, 17(1), 43–54.

16 Called “interlegality”, see de Sousa Santos, B. (1987). Law: A Map of Misreading – Toward a Postmodern Conception of Law. Journal of Law and Society, 14(3), 279–302; de Sousa Santos, B. (2002). Toward a New Legal Common Sense. London: Butterworths.

17 Cortell, A. P. and Davis, J. W. Jr. (2000). Understanding the Domestic Impact of International Norms: A Research Agenda. International Studies Review, 2(1), 65-87; Jordana, J., Levi-Faur, D., and Fernández i Marín, H. (2011). The Global Diffusion of Regulatory Agencies: Channels of Transfer and Stages of Diffusion. Comparative Political Studies, 44(10), 1343-1369; Dolowitz, D., and Marsh, D. (1996). Who Learns What from Whom: A Review of the Policy Transfer Literature. Political Studies, 44(2), 343–357.

18 For simple core theory see Horowitz, M. C. (2010). The Diffusion of Military Power: Causes and Consequences for International Politics. Princeton, NJ: Princeton University Press, pp. 3-4, 9-10, 49. For more theories see Wisdom, J. P., Chor, K. H., Hoagwood, K. E., and Horwitz, S. M. (2014). Innovation Adoption: A Review of Theories and Constructs. Administration and Policy in Mental Health, 41(4), 480–502.

19 For an example of this model see Red Hat Inc., which works in the field of Linux for enterprises (https://www.redhat.com/).

20 Lanier, J. (2010). You Are Not a Gadget: A Manifesto. London and New York: Penguin.