In a recent video interview with Bitcoin Magazine, Troy Cross, a Professor of Philosophy and Humanities at Reed College, discusses his latest piece for Bitcoin Magazine’s “The Mining Issue,” titled “Why the Future of Bitcoin Mining is Distributed.” You can watch the complete discussion here.
During the interview, Troy examines the centralization issues in Bitcoin mining and argues strongly for the decentralization of hashrate. While mega mining operations have emerged due to economies of scale, he emphasizes an important—and potentially economic—need for distributing mining power, providing insights into the future landscape of Bitcoin’s infrastructure.
The following article is part of Bitcoin Magazine’s “The Mining Issue.” Subscribe to get your copy.
Intro
When Donald Trump expressed his desire for all remaining bitcoin to be “MADE IN THE USA!!!” it sent Bitcoin enthusiasts into a frenzy. Mining is beneficial, right? We want it to occur domestically! Indeed, the U.S. is on its path to dominating the sector. Publicly traded U.S. miners currently account for 29% of Bitcoin’s hashrate—an uptick that seems inevitable. Pierre Rochard, the Vice President of Research at Riot Platforms, anticipates that U.S. miners will control 60% of the hashrate by 2028.
However, let’s be realistic: Concentrating most Bitcoin mining operations in the U.S., particularly within large public miners (as opposed to a Bitaxe in every household), is a flawed strategy. If a majority of miners are situated within a single country, especially a wealthy and powerful one like the U.S., miner decisions could be influenced not just by Satoshi’s insightful incentives but also by the political moods of whoever is in charge. If Trump achieves his stated goals, the very essence of Bitcoin as non-state money could be jeopardized.
In what follows, I will describe how a nation-state might attack bitcoin by regulating miners, review the incentives that have concentrated Bitcoin mining in extensive U.S. data centers controlled by a select few companies, and argue that the future of Bitcoin mining will closely resemble its earlier days, marked by a proliferation of miners that are geographically dispersed like the nodes themselves.
Furthermore, I contend that despite some Bitcoiners’ excitement for “hash wars” and political bravado, nation-states have a vested interest in a future where no single country dominates Bitcoin mining. This “non-dominance dynamic” distinguishes bitcoin from other technologies, including weaponry, where the incentives for monopolization compel nations to race to be first. But with Bitcoin mining, dominance equates to loss. Once nation-states grasp this unique game theory, they will act to prevent miner concentration.
The Attack
If the U.S. controlled the majority of hash power, how could bitcoin face an attack?
The U.S. government could issue a directive through the Treasury Department, instructing miners to blacklist transactions from specific addresses, like those in North Korea or Iran. Additionally, the government could prohibit miners from adding blocks to chains containing earlier blocks with disallowed transactions. Major U.S. miners—publicly traded companies—would have no choice but to comply with the law; executives would prefer not to face prison time.
Moreover, even miners located outside the U.S., or private miners in the U.S. defying the law, would still feel pressure to censor transactions. Why? If a rogue miner incorporated a censored transaction into a block, compliant miners would need to orphan that block to build on the previously government-approved blocks. The orphaned block would also deny the rogue miner their rewards, leaving them with no returns for their efforts.
The subsequent events are uncertain, but none of the potential outcomes are favorable. A fork would likely occur. This new fork might adopt a different algorithm, rendering existing ASICs incompatible with the new chain. Alternatively, it could stick with the current algorithm but manually invalidate blocks from known violators. Both outcomes would yield a government-approved bitcoin and a non-compliant bitcoin, with the government-compliant fork maintaining the original code.
When Bitcoiners discuss these scenarios, they typically assert that market participants would abandon “government coin” in favor of “freedom coin.” But would that truly happen? Perhaps we, the individuals reading Bitcoin Magazine, who seek freedom and identify as cypherpunks, would choose to ditch the censored fork for the new freedom variant. Yet I doubt that institutional players like BlackRock, Coinbase, Fidelity, and other Wall Street entities would react similarly. Thus, the relative economic value of both forks, particularly five to ten years from now, remains ambiguous. Even if a non-compliant fork were to persist and retain significant economic value, it would face philosophical and economic challenges.
Now, reimagine the same attack scenario, but with a well-distributed hashrate. Suppose U.S. miners comprise only 25% of the hashrate. In such a situation, if the U.S. government compelled miners to censor certain addresses and orphan new blocks containing those transactions, the consequences would still be dire. However, the 75% of miners beyond U.S. jurisdiction would continue to validate non-compliant transactions, ensuring that the most heavily supported chain would still include those blocks. If a fork happens in this distributed-mining scenario, it would be the government-compliant bitcoin that would need to fork away and abandon proof of work for social consensus.
This scenario remains bleak. U.S.-based custodial services may be mandated to support the new compliant bitcoin, creating an economic threat to the authentic bitcoin, at least temporarily. However, if the mining network endures outside the U.S. and maintains majority hashrate, it could resemble the U.S. opting out of bitcoin rather than seeking to co-opt it through hashrate dominance.
How Did Bitcoin Mining End up in Large U.S. Data Centers?
The evolution of Bitcoin mining serves as a case study in economies of scale.
Let’s rewind to the beginning. In the early days, the functions we associate with miners—collecting transactions into blocks, executing proof of work, and broadcasting their blocks—were all encompassed in Satoshi’s descriptions of node operations. There were no distinct “miners”; every node could mine at the click of a button. Thus, mining was as decentralized as the nodes themselves.
However, CPU mining was swiftly replaced by mining with graphics cards and FPGAs, and then by ASICs starting in 2013. For many years, mining remained a vestige on nodes until Bitcoin Core finally eliminated it entirely in version 0.13.0 of the software in 2016. Once mining developed independently from node operation, utilizing specialized equipment and expertise, it began to scale. This was entirely predictable.
In The Wealth of Nations, Adam Smith describes a pin factory with only ten workers producing 48,000 pins per day, while each employee could only make one pin daily individually. By concentrating on distinct stages of pin production, developing tools for each subtask, and collaborating sequentially, the employees produced significantly more pins with the same labor input. To illustrate, the additional cost of increasing production by just one pin becomes negligible for a factory already producing 48,000, having already invested in equipment and skills; it merely requires a slight increase in labor and materials. Conversely, for an individual making just one pin daily, the marginal cost of adding each additional pin doubles.
Mining, once freed from the limitations of CPUs, featured similar characteristics that lent themselves to scale efficiencies, akin to pin production. ASICs operate as specialized tools, much like pin-making machines. Data centers designed for ASICs’ unique power density and cooling requirements further differentiate this operation. Similarly, mining in a multi-megawatt commercial facility spreads fixed costs across a larger number of mining units, more efficiently than a basement operation. Some relatively scale-independent expenses miners face include:
- Power expertise
- Power equipment
- Control systems expertise
- ASIC repair expertise
- Cooling expertise
- Cooling facilities
- Legal expertise
- Financial expertise
In a larger operation, fixed costs are distributed over more revenue-generating machines. Additionally, scaling from a basement to a commercial facility enhances bargaining power with suppliers and labor. Transitioning from one’s basement to a local commercial space offers better electricity pricing. Moving from an office park to a mega-center allows for hiring power specialists who negotiate sophisticated contracts with energy suppliers and hedge against price fluctuations. Every time a single machine is sent for repair, the average cost per incident rises compared to employing a repair specialist who can fix failing ASICs on-site, assuming the scale is adequate. Furthermore, pricing negotiations with ASIC manufacturers depend on order size, enabling larger companies to secure more favorable terms and squeezing smaller miners just as larger companies do with local retailers.
Economies of scale are unsurprising as they apply, to varying degrees, to virtually all manufactured goods. The benefits of scaling explain how mining evolved from my personal experience with graphics cards in my basement 13 years ago to facilities approaching 1 GW today.
However, scaling is not the sole reason mining has centralized within the U.S. and large public companies. Understanding the latter requires recognizing two additional factors. Firstly, access to financing scales well: Large public firms can generate capital through stock dilution or bond issuance—options unavailable to small-scale miners. While smaller miners can borrow, they face less favorable terms than large firms, and the U.S. has the deepest capital markets globally. Secondly, the U.S. possesses a relatively stable legal framework, minimizing the risk of abrupt state actions like seizing a mining operation or reining in regulations.
A significant factor that attracted mining to the U.S. in recent years was the availability of electrical infrastructure. After China’s ban on Bitcoin mining, it became profitable to mine practically anywhere globally using virtually any ASIC. However, the U.S. had existing power infrastructure, much of which resided in the rust belt, remnants from the exit of U.S. manufacturing to China. Additionally, abundant power in West Texas—stranded wind and solar energy prompted by subsidies but insufficiently interlinked with other regions—also contributed. Following China’s mining prohibition, miners swiftly occupied underutilized facilities in the rust belt, capitalizing on abundant power and inexpensive land to establish data centers in West Texas.
The capability to raise and deploy substantial funds is a significant advantage that compounds with other factors, particularly given Bitcoin mining’s fixed global rewards. With ample market funding, large public Bitcoin miners acquired the newest, most efficient, and powerful ASICs while negotiating favorable power contracts and hiring top firmware and software experts. This not only disadvantaged smaller miners but also allowed larger miners to significantly enhance global hashrate, leading to increased difficulty. When bitcoin prices dropped, miners without the advantages of scale faced diminishing margins, given their debt-laden ASIC fleets. Even a publicly-traded miner in bankruptcy could continue operating its massive arsenal of machines through restructuring, driving out smaller competitors while managing legal proceedings.
Thus, mining expanded from niche hobbyist activities to gigawatt-scale operations, ultimately settling in America. Mining is an intensely competitive commodity industry, and the efficiencies gained from scaling proved instrumental, particularly when funded by debt and dilution.
Why Mining Will Be Distributed and Small-Scale Once Again
Just as there are economies of scale, diseconomies of scale exist, where unit production costs can increase beyond a certain size. For example, it’s evident that a single gigantic food factory cannot feed the entire world every meal. While efficiencies exist in factory-produced food—evident in the rising average farm size over the last century—limits do exist. Fresh ingredients must travel from farms to production sites, and the final products then need distribution to consumers. Both the raw inputs and final outputs of a food factory are perishable and substantial, making shipping costs excessively high, and affecting quality relative to local markets that can supply fresher produce. Similar principles apply to why sawmills and paper mills are located near forests, and bottling plants near freshwater sources.
Nonetheless, transferring bitcoin incurs no costs: It’s merely about making a ledger entry on the Bitcoin blockchain, which takes seconds. While I enjoy boasting about the artisanal bitcoin mined in Portland, truly, there are no local distinctions that change the quality of bitcoin based on its origin. All bitcoin is fundamentally equivalent. This further supports the notion that global bitcoin production should ultimately centralize in the best possible site for mining.
However, there’s a significant issue with concentrating all mining within a single facility: Bitcoin mining is energy-intensive. Presently, it consumes over 1% of the world’s electrical supply. Electricity constitutes the primary operational expense in bitcoin mining, usually accounting for 80% of costs. Unlike bitcoin, electricity fails to travel effectively. Rather, it behaves similarly to food, which quickly perishes and needs specialized infrastructure for transport. When considering electricity, this infrastructure includes wires, transformers, substations, etc.—all elements of an electrical grid.
The transportation of electricity contributes significantly to its overall cost. What is termed “generation” often constitutes only a fraction of total electricity pricing, encompassing “transmission and distribution” fees as well. As advancements in technology and efficiencies continue to lower generation expenses—like those in solar energy—investments in grid infrastructure only seem to increase. Thus, the idea of transporting electricity globally to a single bitcoin operation becomes impractical. Instead, bitcoin facilities should remain situated where energy is produced, thus avoiding additional costs related to transmission and distribution, sending the bitcoin from these facilities at no extra cost. This concept is already emerging, known as positioning a Bitcoin mine “behind the meter.”
Mining companies may amplify their differences, touting firmware, pools, cooling systems, financing, and management teams. Yet at the core of their operations, there is little separating different mining firms from each other: Each product is identical, incurs no shipping costs, and uses the same machinery (ASICs) to convert electricity into bitcoin. Disparities in electricity costs largely determine which miners will thrive and which won’t. During prolonged price stagnation or a gradual increase, only operations with access to the cheapest electricity will persist.
Consequently, the primary rationale for a global distribution of miners in the future aligns as follows: First, Bitcoin mining is, by design, drawn to the cheapest energy worldwide. Second, this cheap energy is inherently distributed across the globe and often “behind the meter.” Therefore, a third conclusion follows: mining will also be geographically dispersed and situated behind the meter.
For argument’s sake, let’s say Trump’s vision materializes: all mining occurs within the U.S., and it reaches an equilibrium with extremely tight margins. If someone discovers power elsewhere that’s less expensive than the average U.S. miner’s rate, and they deploy ASICs in that location, the overall hashrate would increase, causing some U.S. miners (those with higher expenses) to face closure. This cycle would repeat until mining activity is solely situated along the world’s cheapest energy sources.
Certain types of inexpensive energy exist: gas in the Middle East and Russia; hydro projects in Kenya and Paraguay; solar in regions of Australia, Morocco, and Texas. The widespread availability of energy occurs due to nature distributing it. Rainfall patterns and variations in elevation yield rivers throughout the globe. Fossil fuel deposits are abundant in numerous regions. Wind blows in various locations, and the sun shines almost universally.
Indeed, the distribution of energy is somewhat guaranteed by the solar path around the planet. As sunshine reaches its peak intensity, its energy is liable to become wasted by solar-powered systems, as energy infrastructure is not configured for peak production. I envision a future where a substantial segment of the hashrate aligns with this solar path, with machines utilizing excess solar energy during that time or, in the case of older units, activating only during periods of surplus production exceeding grid demands.
This primary argument can be slightly modified to yield other projections regarding mining’s future. I foresee an abundance of affordable power on a smaller scale, while massive scale sources of low-cost energy (100 MW+) will become constrained. Consequently, if Bitcoin mining continues to expand, small-scale mining will see a resurgence, and the trend toward massive facilities will reverse as large-scale cheap energy sources diminish.
To illustrate why cheap power predominantly exists at smaller scales, we can examine individual cases. Specifically, we can investigate why flare-gas waste emerges in a distributed, small-scale manner and why solar inverters frequently operate undersized, resulting in energy clipping across the system. However, I prefer considering the broader underlying principle. Instances where we possess cheap energy at scale often arise from significant missteps. For example, building an unnecessary dam or nuclear plant can be classified as errors. These large-scale mistakes are limited in number, as they come with high costs and present limits on fiat imprudence.
On the other hand, smaller-scale discrepancies in supply and demand are likely more commonplace. For instance, if gas production at an oil field is significant enough, it will warrant constructing a pipeline for transport; otherwise, stranded gas will result. Similarly for landfills: larger facilities often generate electricity and are grid-connected, whereas smaller landfills struggle to even capture methane, let alone convert it into electricity and deliver it to the grid. The same principle applies to dairy farms.
Furthermore, bitcoin isn’t the only energy-intensive computation form. When abundant cheap energy is available, other computational methods will settle into those areas. These alternatives, at present, may not scale as well as bitcoin and tend to be less responsive to electricity pricing, thus driving them to outbid bitcoin miners. This indicates that the era of mining reliant on ultra-cheap, large-scale power is limited. Conversely, if you’re mining bitcoin by mitigating flare gas on a desolate oil patch distant from pipelines, the probability of being outbid by competitors for energy-intensive tasks is virtually non-existent. The same holds true if you’re operating with surplus residential solar. Mining on small-scale wasted energy has less commercial appeal for competitors but remains advantageous for Bitcoin miners. Bitcoin mining can scale down sufficiently to tap into these energy reserves, unlike other energy-consuming processes.
Another iteration of the previous argument revolves around the globally distributed demand for waste heat. All electrical energy entering a bitcoin miner is conserved and exits as low-grade heat. Miners are employing this waste heat for heating greenhouses, communities, and natural hot baths. Nevertheless, heating requirements can typically be fulfilled with a small number of machines. A single ASIC or two can effectively heat a home or pool. Utilizing waste heat for electrical heating lowers overall mining costs. In simpler terms, on average, a miner who sells their heat stream will likely find better financial outcomes than one who chooses not to. Thus, here’s another argument for why mining will be globally distributed and at a smaller scale: The demand for heat is overseen globally—intent on meeting needs often found in the far northern and southern regions—yet it can generally service limited scales.
As previously stated, I believe Bitcoin mining will naturally gravitate towards the world’s cheapest energy. This trend typically holds if bitcoin prices rise gradually. In a vigorous bull market—one we’ve witnessed several times—Bitcoin miners tend to utilize all feasible energy sources available, wherever machines can be plugged in. If bitcoin were to surge to $500,000, all previous analyses would face invalidation. Yet in such a bullish scenario, mining would also become globally distributed, not due to widespread distribution of the cheapest power, but because of the availability of power. Bitcoin priced at $500,000 would make all ASICs profitable under any power terms, leading to a situation where no single country, including the U.S., possesses enough infrastructure to manage such demand shocks even if they wished.
It’s noteworthy that high-margin periods tend to be fleeting, as increased ASIC production will chase profits, eventually driving down margins once more. Hence, over time, Bitcoin miners’ geographical distribution will still be determined by the presence of the world’s cheapest energy.
For my arguments to hold, the diseconomies of scale must outweigh the economies of scale referenced above. Assessing the balance between these two factors requires an in-depth examination of each mining business’s spreadsheets, which is beyond the scope of this article.
In summary, I believe that provided the variance in energy costs is substantial enough, this will outweigh any other considerations. However, I cannot claim to have delivered a definitive proof in these claims. These discussions offer broad insights, while the minutiae remain exercises for further exploration.
Geopolitics
Until now, I have examined miner incentives without considering nation-states. Certain countries are acquiring bitcoin, while others leverage their energy to mine it. Nation-states possess motivations that extend beyond Satoshi’s original vision. For instance, Iran might mine bitcoin to monetize its oil, given that sanctions complicate its sale in open markets. Similarly, Russia could engage in mining for related reasons. Such nation-state actors could afford to “mine at a loss” relative to independent miners paying for their energy due to taxpayer subsidization. Their large-scale mining endeavors may render profitability challenging for other miners, potentially pushing some marginal miners out of the market.
However, I do not foresee nation-state mining ultimately concentrating hashrate. Currently, mining in Russia and Iran can be beneficial for Bitcoin, acting as a counterbalance against dominant U.S. public company interests. Moreover, should a nation-state produce an outsized portion of hashrate, I expect other nations holding stakes in Bitcoin’s success—or large bitcoin holders—would likewise begin mining at a loss to ensure mining remains decentralized.
The underlying game theory isn’t straightforward. Instead of a competitive landscape aiming for dominance, bitcoin operates under a framework where all parties benefit when no single entity holds control, while everyone loses when dominance occurs. Unlike almost all other technologies or weaponry systems, where the optimal strategy focuses on achieving global control, we witness a race for supremacy in sectors like battery technology, chip manufacturing, drones, and AI. This is colloquially known as the “Thucydides trap” in foreign policy, which typically fosters preemptive strikes against emerging rivals—given the massive rewards for premier status and the incalculable loss for becoming second.
Yet, achieving dominance in Bitcoin mining is detrimental for Bitcoin mining in general, and consequently harmful for bitcoin itself and its stakeholders. As Bitcoin mining becomes concentrated within one nation, there arises the potential for attacks on bitcoin’s neutrality, a fundamental aspect of its value proposition. For example, Russia might resort to holding bitcoin to circumvent U.S. actions against its fiat reserves, as witnessed during the Ukrainian crisis. If mining becomes centralized in the U.S., Russia would have legitimate concerns that its addresses could face blacklisting by the U.S. Treasury Department. Should this threat materialize, Russia would likely liquidate its bitcoin in favor of alternative assets. U.S. miners might experience a rise in block rewards by dominating other miners, but the value of those rewards could fall with bitcoin’s price. Thus, U.S. miners would have strong motivations to keep Russia involved in mining and discourage them from offloading their bitcoin. The overall argument infers that U.S. miners shouldn’t aim for a form of victory that results in their own detriment. If bitcoin holds significant economic weight within the U.S., policy would ideally incentivize miners to ensure no single country achieves majority control. Essentially, Bitcoiners should advocate for a scenario in which the U.S. amasses enough bitcoin so that no country, even itself, retains a majority stake. Though this may not serve as an inspiring slogan for a campaign rally, it encapsulates the only logical stance for a Bitcoiner.
Disclaimer: Opinions expressed are solely those of the author and do not necessarily reflect those of BTC Inc or Bitcoin Magazine.
