At the start of 2020 we set out to learn about the environmental impacts of the Light Phone II as they mainly relate to climate change. Below is our full documentation of this process, from carrying out a life cycle assessment to purchasing carbon offsets and considering sustainable product design. If you are interested in a shorter summary of the work we’ve done, visit this link.

table of contents:

1 — Introduction
2. Determining the Light Phone II Carbon Footprint
2.1 Our ‘lightweight’ LCA methodology
2.2 The Light Phone II Carbon Footprint
2.3 Interpreting Results
2.4 There’s more to sustainability than CO2
2.5 Takeaways from the Light Phone II’s carbon footprint
2.6 Takeaways from our emissions research process
3. Carbon Offsetting the Light Phone II
3.1 Why carbon offsets?
3.2 Offsetting via carbon sequestration
3.3 Our offset-friendly checkout
4. Thinking Long Term

—

1 — Introduction

Most writing about climate change tends to come from a place of diagnostics. What is it? How bad is it? Who’s to blame? Sometimes this rhetoric drifts towards solutions. What can we do about it? But even when grand visions of forward progress are laid out, the complexity of these plans can be just as daunting as the problem itself.

A similar thing can be said of how we understand the impact technology is having on our day to day lives. There’s no shortage of stories about the troubling way we use smartphones, or the ever-creeping surveillance arm of large technology corporations. Like with climate change, solutions to these problems tend to take place in an idyllic and distant future. For many the question remains, what can we do now?

As much as it was a real world product, the first Light Phone was an experiment in what we can do now about the dangers of smartphone addiction. While its radical simplicity might not have been the best answer for everyone, it wasn’t a concept phone. Most recently, the Light Phone II addresses some of those practicality issues and takes the idea a step further by actually replacing the need for a smartphone for many of our users.

In the realm of climate action, it’s fair to say that we need more examples of readily available actions that can be taken by individuals, organizations, and governments alike. We need more practicality. We’re increasingly past the point of being able to wait for the perfect source of energy, form of protein, or method of transportation. If this isn’t already clear to you, there are many examples of media and scientific literature that help convey this reality. As such this report won’t waste time reiterating the obvious. Instead we’ll be focussing on three steps we’ve taken to start addressing our phone’s climate impact: 1) assessing its carbon footprint 2) providing a way to offset this footprint and 3) informing smarter product design going forward.

The ‘information communications and technology’ (ICT) sector specifically often goes unnoticed when it comes to its role in climate change, as well as its role in a range of other environmental impacts. The reality is that smartphones, computers, and other electronics devices play an increasingly important role in our lives and as such, require increasing amounts of resources to be produced and used. While only a couple of percentage points of our global emissions can be attributed to the ICT sector at the moment, many believe that this share will disproportionately increase relative to other sectors in coming years.

In the spirit of asking “what can we do now?” the Light team decided to figure out what it would take for the Light Phone II to be carbon neutral. This means negating the carbon (equivalent) emissions of the device over its entire life. We wanted to make this project public and available for others to take inspiration from (you’ll realize if you read through that we’ve taken quite a lot of inspiration from other projects like the Fairphone). Although determining an electronic device’s carbon footprint makes for tricky accounting, ultimately coming up with a value for the phone’s lifecycle emissions is by no means impossible, especially if working off of previous examples of reporting. From here, negating a given amount of emissions is just a matter of paying a few dollars per device in the form of carefully selected carbon offsets. It may only represent a short term solution but it’s affordable, accessible, and actionable.

2. Determining the Light Phone II Carbon Footprint

2.1 Our ‘lightweight’ LCA methodology

For this study, we conducted what can be described as a ‘lightweight’ Life Cycle Analysis (analysis is used interchangeably with ‘assessment’), or LCA that had the explicit goal of determining greenhouse gas emissions for the purposes of purchasing offsets. LCAs are considered the standard method used for understanding a given material processes’ environmental impact, whether that be the creation of a single consumer product or an entire industrial system. The origins of LCAs date back to the postwar period where many began to ask what the underlying impacts of an increasingly industrialized society were. The methodology picked up steam in later decades when it was used as a tool by corporate advertising departments to assert quasi-scientific claims about their product’s environmental impact. Since then, LCAs have turned around and gained acceptance as the best practice for rigorously determining something like embedded carbon emissions. The ISO Standard 14044 shows in the greatest detail what it means for something to be considered a proper LCA.

In practice, an LCA is a sustainability report for a product. A popular format for LCAs breaks down a product into its material components and then attempts to measure each of these components along the lines of environmental performance over the entire product’s life. In all but the most rigorous LCAs secondary data is the main tool used to estimate these environmental impacts based on known averages. The logic for this is that for most products today consist of hundreds of inputs that come from a complex globalized supply chain. Using averages allows researchers to make faster work of understanding complex upstream interactions. For example, in the context of a mobile phone, the standard LCA involves weighing and measuring discrete electronic components — transistors, integrated circuits, a battery, plastic casing, etc — and then using an LCA database to calculate net impacts based on the industry averages for each of these categories. If this seems largely like guesswork, that’s because it is, but as long as one uses high quality models, this approach can be a sufficiently accurate substitute for the impractical process of gathering only primary data, i.e. surveying every factory behind every screw and asking for sensitive information like their inputs and energy usage.

Past being an estimation heavy method, there remains well documented problems with LCAs, namely their inability to be an accessible and a cost/time effective approach for those that sit outside of academia and corporate R&D. Let’s say you are a small to mid sized electronics manufacturer like us — conducting an LCA just for carbon emissions on a single phone model might typically set you back north of $15k in consulting fees. Just the proper license to LCA software and an accompanying dataset can cost upwards of ten thousand dollars.

This cost is especially steep considering that it provides a service that is unnecessarily complex for our end goal, which is once again, to not just understand the Light Phone II’s climate impact, but to offset this impact. So ultimately if you could imagine a level of rigor that is somewhere between cold guessing and spending lots of time scrutinizing over every step of the steps of the Light Phone’s production, we were looking for something in the middle, hence a ‘Lightweight LCA.’ All together our two objectives were to (1) simplify calculations as much as possible without creating an inaccurately low estimation (2) lower the logistical and financial barriers to performing a LCA.

In order to achieve both these goals, our starting point was recognizing that even the most sophisticated LCAs still tend to rely on the process of component based estimations. This means that as long as a set of accurate coefficients for carbon emissions (equivalent) could be ascertained from preexisting LCA studies, it would be possible to circumvent the need for purchasing an entire LCA database, all while conducting estimations that were comparably accurate to more involved LCA studies. By way of example, if you can parse a recent mobile phone LCA and see that 10kg of CO2 are attributable to 10 cm2 of integrated circuit die area, then you also can assume that this LCA was operating off a coefficient of 1kg CO2 per 1cm2 of IC die area. Because this carbon intensity coefficient represents a generalized estimation of all the upstream emissions attached to the production of IC wafer for an electronic device, then it’s possible to extract this coefficient out to other devices assuming that the geographical boundaries of the two devices are similar. This coefficient extraction process can be replicated with each component category across a given device.

Of course this calculation is more complicated in reality. Namely, there can be very significant differences between the same type of components made in the same location. An IC produced in 2010 looks very different from one produced in 2020. So effort needs to be put into making sure that if there is potential for error, this error is only one of overestimation. Since our main goal is to completely offset the device’s emissions, if we purchase extra offsets, the only harm this produces is helping mitigate climate change a little bit more.

The other main difficulty with this strategy is that it only gets you as far as there exists granular data for each component you want to measure. In a mobile phone that has hundreds of different types of components, producing estimations for each of these categories is not only impractical but simply not possible due to the limited coverage of LCA databases. The method of working around this complexity is to first focus on electronic components like the display unit or CPU that are known to have the greatest environmental impacts. From here you can run a regression analysis on a population of devices that have primary LCA data for all of its components and see which components have the strongest ability to predict the impact of other components. Put simply, this backtesting method can allow you to reasonably estimate the environmental impact of many miscellaneous components of a phone by just knowing something like the silicon die area within integrated circuits (see section 2.6.1. here).

For our ‘lightweight’ LCA this is exactly the method we leveraged to help us make reasonably accurate estimations without having to survey hundreds of upstream suppliers halfway across the world. By simply using a Bill of Materials that includes all of the Light Phone II’s components and their weights, we can arrive at a conservative estimate for the phone’s GHG emissions.

There are other phases of a mobile phone’s lifecycle besides its production. Raw materials need to be extracted as inputs for production, the device is shipped to the customer, the customer goes through hundreds of charge cycles, and the device is finally disposed of in some manner (‘end of life’). To calculate the climate impact of these other phases we relied on a similar blend of primary and secondary data.

We captured all of these phases, as well as their relevant primary and secondary data in this table here (also embedded below). See the columns labeled ‘Formula Explanation’ and ‘Accuracy Commentary’ to learn more about the justifications for each LCA calculation. When possible we included an input-based formula for each CO2 calculation (see ‘Light Phone Input’ and ‘CO2 Coefficient’ columns) but for more complex formulas we simply listed the final CO2 count and left this column set to ‘1’. Included in the table are the sources for secondary data. The source most worth mentioning here for its usefulness was the research done by Orange Mobile (Andrae et. al.) and the nonprofit Forum for the Future, in which they developed a streamlined methodology for mobile phone LCA called Open Eco Rating (v3), which captured basically the LCA aggregation process we aimed to achieve on our own. We used the Open Eco Rating tool as a foundation for many of our high level LCA calculations, and made augmentations based on the availability of better data. Given two reasonably accurate methods of estimation we sided on the method that would yield more emissions, or a worst case scenario. The substitutions from Open Eco Rating data came from other phone-specific LCAs such as the Fairphone 2 LCA and the various Sony Ericsson LCAs.

A standard practice in LCAs is to set a functional unit — or the usage parameters that one might base an inquiry around. For our research we defined the functional unit as a single Light Phone II model used for 3 years. In determining the boundaries for impact categories to analyze we based our study off of other mobile phone LCAs for the sake of comparability. For the production phase, every single material input that is used to make a Light Phone II is accounted for, as well as key nonmaterial inputs such as research and development flights made by team members from the US to Taiwan. In the transportation phase we based our model off a customer receiving a phone by mail based in NYC. For the usage phase we accounted for the electricity consumed by an average Light Phone user recharging their phone every other day for three years using the US electricity grid. We also accounted for essential phone operating system downloads, but not for carrier data usage. Finally in the end of life phase, we based our model on a 100% recycling rate, which is admittedly not realistic but for the purposes of calculating GHG emissions led to the most conservative scenario.

Methodology summarized

• We set out to conduct a ‘lightweight’ Life Cycle Analysis that would estimate the climate impact of the Light Phone II over its entire lifecycle.
• We wanted a result that was still accurate compared to a full fledged LCA, but that could be arrived out 1) without gathering extensive primary data and 2) without spending money on expensive LCA software or datasets.
• We did this by extracting out secondary impact data from preexisting mobile phone LCAs, and plugging in primary data for the Light Phone II.
• If we were to let in error in our estimations we wanted to be sure that this was an error of overestimating emissions, not underestimating emissions.

2.2 The Light Phone II Carbon Footprint

Our study estimated 60.68 kg of CO2 (equivalent) attributed to the Light Phone II over its entire life. By a significant margin the production phase of the phone proved to be most carbon intensive at 55.79 kg, or a total of 91.9% of the entire emissions. Next in line is the transportation phase which yielded 3.50 kg. The usage phase yielded 1.35 kg while the end of life phase yielded 0.33 kg.

Looking further into the production phase it’s clear that the single largest contributor to the Light Phone’s carbon footprint is the manufacturing of integrated circuits (ICs) and other electrical components throughout the phone, which led to an estimated 20.8 kg of CO2. Not far behind at 16.12 kg of CO2 is the production of the display unit (more on this below). Just these two components together account for 60% of the entire phone’s carbon footprint.

2.3 Interpreting Results

Key to understanding the climate impact of the Light Phone is contextualizing the results against similar products. Read enough LCAs and 61 kgs of CO2 might be intuitive but for most people this number is just a construct. How could a 78 gram ‘Light Phone’ produce nearly 800 times its weight in emissions? To help understand our results let’s compare the emissions of the Light Phone II to other mobile phones and comparable electronics devices.

The first thing to note about comparing a set of products along the lines of LCA results like carbon emissions is that this comparison is rarely apples to apples. Although LCA researchers strive to set standards and benchmarks for something like a carbon footprint analysis, in practice there can be pretty large disparities between studies. For instance, this study of the Sony Xperia T makes calculations that include the cellular network activities for the phone, whereas many other LCAs (including ours) do not, even though all studies follow the same basic guidelines. All together this makes pure comparisons between any two LCAs difficult.

That being said, there are general trends that can be derived from a given study’s results. In the case of the Light Phone II, the most interesting question to be asked is, ‘does a phone stripped of non essential functionality lead to a decrease in CO2 emissions?’ The comparison between the Light Phone and a modern iPhone might be most insightful here. The current top of the line iPhone model, the 11 Pro Max, is a full featured smartphone with multiple cameras, a 6.5 inch oled screen, and a 3,969 mAh battery. Compared to the Light Phone II with no camera, a 2.8 inch e-ink screen, and a 980 mAh battery, it’s a very different kind of device. Despite the vastly larger economies of scale that Apple can use to improve sustainability the 11 Pro Max still is linked with 86 kg of carbon over its life. This increase compared to the Light Phone’s 61 kgs is directly attributable to the extra material requirements that it takes to produce the iPhone, as well as the greatly increased energy it will consume over its lifetime. It’s also reflected in the percentage of emissions coming from the different life cycle phases. 78% and 18% of the iPhone’s emissions come from the production and usage phases respectively, whereas for the Light Phone this ratio is much more severe at 91% and 2% respectively.

Past arriving at a general ‘more features more carbon’ logic, this LCA and its lightweight methodology is not well suited to definitively claim that the Light Phone II is the most carbon-friendly mobile device, even though this could potentially be true. There are two reasons for this and they both have to do with the scope of this study. For starters, in some of the most important impact categories (i.e. the production of the display and the IC/electrical components) we intentionally made conservative estimations that would likely make for inflated emissions levels. As an illustration, one of the most fundamental design features of the Light Phone is its e-ink screen, but there is no available data (paid or public) about the carbon intensity of e-ink display manufacturing. So we were forced to use an estimation based on LCD displays, which are likely more carbon intensive. Secondly, this LCA was different from your typical LCA which sometimes has an implicit bias towards producing as low of a result as possible (manufacturers want to claim their devices are more environmentally friendly). Because our ultimate goal was to offset emissions and to improve the state of climate change, an upwards error (and thus increase in purchased offsets) was actually preferable to us.

Ultimately, we decided to conduct a study of the Light Phone II carbon footprint after the final blueprints for the phone were submitted to the manufacturer and tens of thousands of phones had already been shipped to users. Instead of conducting an LCA that would have something to say about the particularities of a production process that was already well in motion, we were more concerned about how to fix the damage that had been done. Looking forward to future Light Phones, we’ll be able to use the broad strokes of this study to guide a more proactive sustainability strategy that can be concerned with the fine details. For better or for worse, leaning too far into these details would actually make for a less desirable result, which is a decrease in purchased offsets.

2.4 There’s more to sustainability than CO2

If you’re any bit familiar with the problem of e-waste, you know that carbon emissions are just one area of social and environmental concern when it comes to consumer electronics. Of equal if not greater importance to greenhouse gasses are the effects that electronics have when it comes to the mining and depletion of rare earth metals, or the completely broken system of disposal that comes after someone is done using a device.

Although we didn’t focus on these other issues for this study there are still takeaways about the Light Phone II that deserve to be mentioned. When it comes to rare earth metals, the story is similar to many other electronics devices. Our battery uses cobalt and there are traces of gold and other conflict minerals throughout the phone. Because of our lack of proximity to the upstream producers of the phone’s components we are not aware of what quantities of these elements are used, nor where they were sourced from. Figuring this out is very difficult but not impossible. It just requires more time and resources, which might be available in future Light sustainability projects.

Speaking to the longevity of the Light Phone II, we don’t have data around how long users typically keep their devices, so we used the standard number of three years. In a rapidly evolving consumer electronics sector where each year comes with a flurry of new phone models, we think the Light Phone can embody a device that can realistically be used for at least three years. But one thing that stands in the way of this is repairability. We’ll admit outright that in pursuing a phone with a clean minimalistic design, repairability was not something we prioritized. We’re starting to see Europe regulate repairability in the electronics sector, something we hope becomes more mainstream in coming years.

When it comes to the end of the life phase for the Light Phone II, our LCA model is based off of a 100% recycling rate (for carbon calculation purposes) but this rate in reality in the US is closer to 20%. Even though we strongly encourage all Light Phone users to not throw their phones in the trash and instead find a trustworthy electronics recycling program, we have no official recycling or trade in program. This is something that should receive more attention in the future, but there might always be economic barriers in place around creating such a program. Innovations such as Apple’s Daisy phone disassembly robot only are possible when working at very large scales.

Just because we decided to focus on climate change does not mean that other types of externalities are of secondary importance. Matter of fact, the more time that was spent learning about these issues the more it became clear that they are all interconnected. Here we recognize and applaud the Fairphone team for being clear leaders around approaching electronics sustainability from a diverse perspective. As mentioned above, future LCAs that we conduct will take time to consider the impact of these other categories, and future Light Phones will aim to better mitigate their impacts.

2.5 Takeaways from the Light Phone II’s carbon footprint

• An E-Ink display means less electricity day to day: Even though we weren’t able to find secondary data around the carbon emission of producing e-ink displays, we do know from our analysis of the Light Phone’s power consumption that it uses about a quarter of the power of full featured phones. This is in large part due to the energy efficient nature of e-ink displays. Over the lifetime of a phone this can lead to some pretty sizable differences for emissions associated with daily charging. Nonetheless it would still be valuable to quantify any potential differences in embedded emissions that come from display production, as currently this is one of the most carbon intensive aspects of the phone.
• Less features means less components: Which means fewer GHG emissions. Let’s say that the Light Phone had a 12mp rear camera, a 20% bigger screen, and twice the amount of storage. This could result in an total emissions footprint increase of as much as 30%. A phone that cuts out unnecessary componentry is able to make significant reductions in its emissions levels.
• High impact improvements over greenwashing: If you were a consumer with a casual interest in sustainability and saw that the Light Phone III was ‘Now Made with 100% Recycled Plastic!©’ you might be impressed by this. But analysis in the emissions composition for electronics reveals that such an improvement would have very little effect on the overall footprint for the device. As such, it might be a better use of time to understand what gains there are to be made around reducing the carbon intensity of something like IC production, or designing higher performance software to run on a simpler chipset.
• R&D travel isn’t negligible: One example of an easy improvement that could be made to the Light Phone II carbon footprint is to reduce the number of flights taken by the Light team to Taiwan to help guide the phone’s production. All together there were over 10 flights for multiple team members, totaling nearly 4kg of CO2 (eq.).
• Primary vs companion phone: A major consideration for the sustainability of the Light Phone is whether or not it’s used as a primary or secondary device. For many Light Phone users, the device is used as a backup or companion phone to a more full-featured smartphone. What this means is that even if the Light Phone is overall more sustainable than the iPhone 11 Max Pro, if the Light Phone doesn’t act as a complete replacement, then these sustainability gains are unrealized. Matter of fact, one could argue that the Light Phone is existing where there was previously nothing, which in LCA terms means that it suffers from additionality. In an ideal world we want the Light Phone to be the only phone that our users own. It’s connected to our philosophy of stripping down our digital tools to only what’s necessary. With the Light Phone II we hoped to make a device where this is more possible than ever. With new features and firmware coming soon this will be even easier but there’s still room for improvement on this front and we hope that future Light Phone models make it easy to just own one device.

2.6 Takeaways from our emissions research process

• Breaking the sustainability analysis cycle: Now that we understand better what plays the largest role in the Light Phone’s sustainability credentials we are able to integrate this knowledge into the next product development cycle, which in turn helps increase the availability of data that can be used for future LCA projects. Conversely if we never took the time to create this initial understanding it would be easy to carry on with future products without an eye for sustainability, which would make future LCA projects more difficult. A lesson we have for other smaller product companies — don’t let perfect be the enemy of good when it comes to taking on sustainability initiatives. Just because Apple’s sustainability budget is probably in the order of billions doesn’t mean that there aren’t real and practical steps that your company can take today using far less resources. The first step might be the hardest but once you get the ball rolling, the logic of sustainability becomes woven into how your company does business.
• The positives and negatives of conducting an inexpensive LCA: As mentioned above, one of the major constraints of this project was resources. It was mainly carried out by a single person with a budget of only a couple thousand dollars. Most notably this meant circumventing purchasing expensive LCA datasets and software. Just the fact that you are reading this post means that this approach, which began as an experiment, was somewhat successful. But it’s not without serious drawbacks that might make its applicability to other product categories difficult. Namely, we relied on a fairly extensive body of mobile-phone specific LCAs to base our methodology off of. Worth highlighting here was incredibly helpful Open Eco Rating (discussed in more detail above) which provided a solid starting point for conducting analysis. Without the Open Eco Rating tool, it’s not clear that our lightweight approach would have been possible. This is all to say, for other companies thinking about carrying out their own inexpensive sustainability studies, we recommend that they base the feasibility of this work around the availability of similar investigations from the academic or private sector. Unfortunately the LCA landscape is still fairly undeveloped, and so for many other product categories it might be the case that paying a traditional sustainability consultant is the only option.
• Moving past the ‘smoking gun’: One of the observations of a mobile phone’s carbon footprint that became clear early on in our investigation is that the footprint breakdown between two modern devices is more similar than it is different. Like with the Light Phone II, in virtually every LCA that we came across, the production of integrated circuits was the most carbon intensive aspect of the device. This likeness is after all what makes the secondary data estimation approach that is used in this report and other LCAs possible. Confronting these similarities is not very exciting though for an investigator hoping to reveal a smoking gun within a phone’s footprint. In our opinion, the implication that this has for sustainability reporting is that more time should be spent researching primary data for the components in a device you suspect breaks free from norms. For us, this might have meant spending more time analyzing the impact of the Light Phone II e-ink display.
• Going even lighter: A point that stuck out to us throughout the research process, at least as it applies to mobile phones, was that tremendous amounts of time can be spent calculating a number that might have the implications on the order of a few hundred grams of CO2. When you further consider that this fraction of a kilogram effectively costs (at the time of writing) less than a cent to offset, it places in question whether or not the time that goes into that calculation is worth it. Now if you are completing a full bodied LCA where other impact areas outside of climate change were being considered, then maybe thoroughly researching each component would be worthwhile. For the sake of our study, where the goal was offsetting first and foremost, in hindsight it seems like we could have gone even lighter with our calculation methodology. At the very least we could have more aggressively estimated the impacts of components with relatively small emissions contributions, and spent this time instead on categories with high impacts. Putting this all together, we actually think that it would be interesting to consider abandoning the component based LCA process altogether in exchange for more simplistic estimation practices. For instance — what if we just used a conservative average for a device’s footprint that was based on a proxy like screen size or processor speed and storage size? What if there was a flat emissions rate, somewhere around 75kg’s of CO2 equiv. that reflected the average of the entire current field of mobile phones? Of course this standardized approach would yield the subtle differences between each phone irrelevant, but if it meant attaching an easy to digest, easy to offset, figure to devices that otherwise wouldn’t be researched, then this might be a strategy to consider.

3. Carbon Offsetting the Light Phone II

3.1 Why carbon offsets?

As if the process of determining the Light Phone II’s carbon footprint wasn’t tricky enough, armed with an estimate for embedded emissions we were then tasked with doing something about it, which in the short term means purchasing carbon offsets. If you’re any bit familiar with the world of carbon offsets, you would know that the ‘offsetting’ part is not always a straightforward thing.

Before we dive into why, here’s a basic definition of what carbon offsets are. Carbon offsets are anything that can measurably decrease the amount of CO2 (equiv.) in the atmosphere. The ‘offset’ part of its name comes from how they are used to counterbalance an action that leads to the emission of greenhouse gasses. A common type of offset is the creation of renewable energy sources. Entity A emits carbon, entity B sells A offsets certifying that they created a proportional amount of carbon free energy to A’s carbon footprint. As you might expect, in practice this equation is not so simple. As such, offsetting has grown into somewhat of a controversial practice in the last decade.

To be clear: our position is that buying offsets should not not serve as a license to pollute. Offsets will not lead to any meaningful improvement in climate change if those who utilize them aren’t also driving down their emissions. They should be considered as just one small step in a climate action plan. Secondly, not all offsets are created equal. In order for a given offset to do its intended job, it needs to account for these criteria:

1) Additionality — Would an offset exist otherwise if it were not for the offsetter’s purchase?

2) Permanence — Will an offset’s effects be guaranteed (effectively) forever?

3) Double Counting — Is there only one entity claiming a (unit of) offset as theirs?

4) Leakage — Does the existence of an offset lead to secondary negative environmental or social externalities that wouldn’t have occurred otherwise? For ex — does protecting a plot of forest lead to more logging somewhere else?

5) Geography (optional) — All things being held equal, an offset produced closer to a source of pollution is preferable to one that isn’t.

Ensuring all these criteria are met is no small feat. It requires fairly rigorous accounting and monitoring that many offsets providers are not well incentivized to carry out. Not to mention the fact that for most of these criteria there is a lot of grey area. Standards and certifications exist for offsets, but in general the space is disjointed and confusing for newcomers. This is somewhat predictable considering the consumer offset market is completely voluntary and has no regulations that guide its practice. Despite these challenges, we thought that choosing high impact offsets was still very much possible.

3.2 Offsetting via agricultural carbon removal

Our search led us in the direction of a carbon sequestration method colloquially referred to as ‘carbon farming.’ One particular project, Hudson Carbon, was on our radar as a compelling spin on this carbon removal method popularized by startups Nori and Indigo Ag. In this section we’ll break down the specifics of carbon farming and the Hudson Carbon project. Feel free to skip to the next section if you just are interested about the end result of our deliberation. Vice versa, there is only so much detail we have space to go into here. Reach out if you have any additional questions.

The basic idea behind carbon farming is to modify farming methods in a way that leads to a measurable increase in the amount of carbon stored in the soil. All plants store carbon naturally as they carry out photosynthesis. This carbon comes from CO2 in the air surrounding them, the same CO2 that contributes to climate change if too concentrated. Compared to the popular method of planting trees which emphasizes carbon storage in the actual biomass (i.e. roots, trunk, leaves) of the trees themselves, carbon farming emphasizes carbon stored in the soil the plant rests in, something called Soil Organic Carbon (SOC).

Carbon farming focuses on SOC because it’s something that farmers actually have a lot of control over. While carbon farming methods can be quite diverse, they are commonly summed up as ‘no till’ or ‘regenerative’ farming. Tilling, or digging up a plot after it has produced its harvest and before a new harvest is planted, is associated with the release of carbon stored in the soil back into the atmosphere. When farmers devise crop planting methods that don’t require tilling, such as through planting cover crops, they are able to keep more carbon in the soil. If you’ve read Masanobu Fukuoka’s classic, The One Straw Revolution, many of these techniques will sound familiar to you.

The key development that made carbon farming feasible as a method for retail carbon offsetting was when scientists developed models that measured the amount of carbon regenerative farming methods could sequester. With the use of a model like the USDA/Colorado State’s COMET, farmers can plug in information about their regenerative practices and get an estimate of how much carbon they have stored compared to a conventional farming baseline. With this information in hand they are able to sell a carbon offset representing this amount, the same way they would sell any other crop. Hence carbon farming.

We chose carbon farming, and carbon farming by Hudson Carbon specifically, because we first and foremost think that the project satisfies the requirements of a good carbon offset. Instead of being purchased through an opaque middle man marketplace, we personally know, and are transacting directly with the Hudson Carbon team. We recently had a chance to interview them regarding the five offset criteria listed above. Without diving unnecessarily deep, we’ll explain the specifics of the project and why we think it’s a good fit. An important thing to note about our justifications for the five criteria is that we are talking about the Hudson Carbon project at the Stone House farm specifically, not all of carbon farming.

Hudson Carbon is a fledgling startup supported by the work of nonprofit, Scenic Hudson Soil Laboratory. They do their work, and are offering their first carbon offsets from a farm plot called Stone House Farm in Columbia County, New York. The origins of this plot are from the Peggy McGrath Rockefeller Foundation wherein it was established as a permanent agricultural easement that could only practice regenerative farming. This is all to say, while Hudson Carbon (and all carbon farming companies) are a relatively recent phenomenon, the project’s nonprofit and philanthropic origins give us confidence that even if the company Hudson Carbon ceases to exist, the offsets we are purchasing will still be maintained. Thus our permanence requirement is fulfilled.

To further support permanence, Hudson Carbon certifies their offsets via the larger carbon farming marketplace, Nori. This means that the Hudson Carbon plots get access to Nori’s buffer pool, which is essentially built-in insurance for the event of a given offset’s failure.

As for additionally, the project’s philanthropic roots illustrate that there isn’t currently a market in this geography for farmers to profitably practice regenerative agricultural methods. To convert from a traditional farm to a regenerative farm costs upwards of three million dollars, and so making this leap is not something that farmers are generally doing without an extra incentive like being able to sell offsets. Furthermore, when the Stone House Farm initially changed its methods to regenerative farming, it was explicitly under the premise of being able to sell offsets with Hudson Carbon.

Leakage, as with any other carbon offset project, is a difficult thing to measure. Still we don’t have reasonable cause for concern around leakage being an issue around our Hudson Carbon plot. The history of this plot as a productive, yet non-industrialized farm means that when Hudson Carbon began practicing regenerative practices there was little to no risk of pushing this industrialized farming to another location.

There is virtually no doubt in our minds that this project has potential to be double counted. Whether through the trust that comes from working personally with the Hudson Carbon team, or the trustlessness that their underlying technology enables in terms of assigning plots, we are confident that the offsets we are purchasing are not being purchased by anyone else.

Speaking towards geography, a big plus, although not necessarily required, is that this project is located so close to Light’s HQ in New York City. This means that in half a day we could drive up and see our offsets working real time, in person. It’s just another feature of the project that gives us confidence in our ability to ensure it’s actually doing its intended job.

Finally, a broader note on the trajectory of carbon farming and how Hudson Carbon fits into this. While not a requirement of a good offset, something that makes us excited about this project are its implications for the broader practice of carbon farming, and carbon farming’s role in combating climate change. Hudson Carbon’s novel approach to measurement that goes beyond the basic requirements imposed by the COMET model means that it has potential down the road to establish a new model that more accurately quantifies carbon storage, and reward farmers for it. At the moment, it’s often the case that carbon farming in non-philanthropic contexts doesn’t make economic sense. But if better models could illustrate more value in the practice than what is currently accounted for, it might make carbon farming a more viable product for farmers. In an agricultural landscape defined by monocultures and global volatility, having farmers see the economic sense in converting their land to regenerative agriculture could spell big wins for not only climate change, but for the other environmental impacts as well.

After vetting Hudson Carbon we agreed to purchase offsets from them for $100 a ton. This is intentionally much higher than the market rate for offsets purchased through Nori, which is approximately $7 per ton. There are multiple reasons for this. For one, as mentioned above the economics of novel carbon sequestration methods often don’t make sense. As Stripe has illustrated on a much larger scale than us — offering to purchase offsets at intentionally high prices allows one to support novel sequestration projects. As these projects receive more support, they are able to lower their costs to levels where more can purchase, and the process repeats itself. The second reason why we decided to buy offsets at that price is because we simply aren’t offsetting a lot of carbon in the scheme of things. At 61 kgs of CO2 the Light Phone II will not be any users largest carbon budget expense, and cumulatively we don’t sell millions of phones like Apple might. So spending more on offsets helps us generate enough revenue for Hudson Carbon to actually make an impact. Finally, we decided to purchase offsets at $100/ton because of the tricky psychology of offsetting. When offsets are too cheap it incentives ‘pollute-and-forget’ behavior. Who really cares about the climate impact of a purchase if offsetting its impact costs only a few cents. To actually adjust the psychology of consumption to account for externalities we need to be able to feel the burden of these externalities. At $100/ton offsetting the Light Phone II isn’t prohibitively expensive, but expensive enough to make you think about what you are purchasing. As with everything else we do at Light, evoking conscious thought is core to the experience.

3.3 Our offsets at point of sale

Now that we had the right offsets at the right price, we were simply left with the task of letting you purchase them. Starting now, as you check out with a new Light Phone on our online store, users are automatically brought to a standalone page that allows them to add on an offset to their purchase for $3.05. This cost is actually half of the total cost to offset 61 kgs of carbon at $100/ton because Light is splitting the cost with you. Just as much as we think it’s the user’s responsibility to act on their carbon footprint, we have to embrace our company’s role in enabling it.

For users that already own a Light Phone, you too can purchase offsets by visiting this dedicated link.

Will will periodically be sending payments in batches to Hudson Carbon starting now. In the future our software team will be building out an offset checkout API that will allow for offset purchases to be automatically routed to Hudson Carbon and assigned to a unique 3ft. x 3ft. plot at the Stone House Farm. If they so desire, users will be able to individually track, and maybe even visit if they are located nearby. Keep an eye out for when we announce this feature.

4. Thinking Long Term

As it has been mentioned throughout this report, our goal for understanding the climate impact of the Light Phone was not to just purchase offsets and then call it a day. Not only would this be an insufficient strategy for combating climate change, it also would be misaligned with the philosophy of our company. The Light Phone represents respite from more features, bigger screens, sharper cameras, infinite content. In creating this space it allows us to reconnect with the more meaningful parts of our lives. The key distinction here is that using a Light Phone is not about pushing away life; it’s about engaging with it. We think that an important part of engaging with life is confronting our impact on the world. Accordingly, something wouldn’t feel right if ‘going light’ meant passively turning a blind eye to climate change. Equally, if our approach to confronting the Light Phone’s climate impact was to pollute, pay, and forget, this too would be untrue to what Light represents.

As such, we wanted to lay out some future design improvements uncovered by our investigation that have potential to actually reduce the Light Phone’s carbon footprint and the general sustainability credentials for the phone. Here are a few that are the most actionable from our perspective:

Lower the bar for using the Light Phone as a primary device — Recently we released the first of several applications that we intend to expand the functionality of the device without disrupting the core Light user experience. A worst case scenario for us is if someone purchases a light phone, finds it unsatisfactory, and never uses it. That’s a lot of resources that were used just to send a device to the landfill. A best case scenario here would be if someone who has previously used a smartphone to the end of its life decides to get a Light Phone and no other phone as their replacement device. Similarly, a middle-of-the-road scenario would be if this user keeps their old but still-functional smartphone as a backup device for when they really need a full featured device, and integrates the Light Phone as their ‘daily driver’. To summarize, we hope to accomplish this in the Light Phone II with continued firmware patches.

Building more performant software — The software running on a device determines exactly how much of the resources are used at any given time; this roughly corresponds to how quickly the battery drains. Some of the deeper firmware implementations can have the biggest effect here, for example, the antenna tuning algorithm. When the device is out of range or in low service areas — it goes to work trying to find a signal. There is certainly room to improve the effectiveness of that firmware here — ultimately resulting in a longer lasting battery, and less energy used over the lifetime of the device.

Improve the repairability of future models — There is unfortunately little we can do at this point when it comes to improving the sustainability of the Light Phone II’s hardware. One of the key points our analysis above has revealed is that repairability is a major factor in improving the longevity of a device. It’s something that electronics manufacturers have not been well incentivized to prioritize when it comes to product design. Two notable examples of this are hard disk storage and battery life. Let’s say that a user fills the Light Phone’s 8gb flash memory, or uses the device frequently enough where their battery health has severely diminished. With the Light Phone II it is very difficult, if not practically impossible, to open up the phone’s plastic casing to make these repairs and/or upgrades. As the Fairphone team has shown us, there’s lots of room for improvement when it comes to making repairability a priority in a mobile phone. The minor cost that comes with making these improvements is generally the portability and streamlined nature of a device. In future Light Phones we look forward to finding a compromise between keeping the phone light and compact, and allowing a user to make repairs on their own.

Create end of life guidelines — This one is straightforward enough. Right now we aren’t clear about what we recommend users do when their Light Phone reaches the end of its life. If users follow their own premonitions then it’s likely that they will throw the phone in the back of the closet, telling themselves that they will use or sell it someday, but never actually get around to it. Ultimately when cleaning their closet they might think to recycle the device, but more than likely it will find its way to the landfill. This is most problematic not for climate change but for the toxins that leach out of the phone if it is not properly disposed of. So, we are currently working on creating a formal end of life recommendation for our users that will likely consist of a recycling program that they can send their phone (for free of course).

Carry out a more detailed LCA for future products earlier in the design process — One of the takeaways from our LCA methodology above was that you can’t really ever know the exact level of emissions that are attached to the usage of a product. This isn’t necessarily problematic if you are taking a conservative approach, and focusing on offsetting carbon after the fact. But what if you’re trying to make changes between one product generation to another? If you are relying on an estimation-heavy method of LCA reporting, then the work that goes into making a product more sustainable will never quite be accurately picked up. Because of this we know that future sustainability analysis from us will likely be a more time and resource intensive endeavor, but one that will help us integrate previous learnings into future product development lifecycles.

Improve the way we talk to our users about consumption and sustainability — Finally, a somewhat meta improvement to sustainability at Light. As hopefully illustrated by this report, we want to do a better job speaking to our community about sustainability. Crucially, we don’t want to use this topic as a marketing tool where a certification badge here and there helps sell more phones. When done right we think that sustainability is a back and forth conversation between a company and its users. After all, if there’s anything our Life Cycle Analysis has proved, it’s that the production phase for the Light Phone is just one of several contributors to its cumulative environmental impact. The usage and end of life phases fall largely on our user’s shoulders. As such we think it’s important to engage our users in their role in sustainability concerns. An example of this was our decision to make the phone’s carbon offsets an optional add-on instead of including this price into the total cost of the phone. Light matches the offset cost 50/50. We hope that you can join us in seeing the importance of treating these issues as more than a concern on the side, and instead a fundamental part of what it means to own and use a Light Phone.

Thanks for reading,

The Light Team

—

Acknowledgements:
Special thanks to Bryan Lehrer for conducting this LCA for Light.
We’d also like to thank Sanctuary Computer for sponsoring this report.