Ethereum's Switch to Proof of Stake Left Apps Intact, But Developers Had Homework to Do
Ethereum completed its transition from proof-of-work to proof-of-stake on September 15, 2022, cutting the network's energy consumption by roughly 99.95% without disrupting wallets, smart contracts, or decentralized applications. For most users, nothing changed. For developers, several targeted but consequential changes required attention.
The technical groundwork was laid in a developer guide the Ethereum Foundation published in November 2021, nearly a year before the upgrade went live.
The upgrade, known as The Merge, joined Ethereum's original execution layer (the transaction-processing mainnet) with the Beacon Chain, a separate proof-of-stake system that had been running independently since December 2020. The result was a single network where validators replace miners as the mechanism for confirming blocks. Block times tightened slightly, from a proof-of-work average of around 13 seconds to a consistent 12 seconds per slot. More significantly, the network gained cryptoeconomic finality: after roughly 12.8 minutes (two epochs of 32 slots each), a block is considered irreversible unless an attacker is willing to have more than 3 million ETH slashed, or destroyed, as a penalty.
The Ethereum Foundation was explicit that The Merge did not address gas fees or throughput, and that smart contract behavior remained unchanged. Fee and throughput improvements were deferred to later upgrades involving rollups and data sharding. "The Merge did not change anything for holders or users," the foundation noted in its documentation. Existing applications continued functioning without migration.
Key Changes That Required Developer Attention
Despite the overall continuity, one technical change created a real risk for smart contract developers. An update called EIP-4399 repurposed the DIFFICULTY opcode, which previously returned a mining difficulty value, to instead return a randomness value from the Beacon Chain called PREVRANDAO. Any contract that used block.difficulty as a source of randomness was either broken or unsafe after The Merge, and the two failure modes are meaningfully different. Contracts that assumed the old numerical range of the difficulty value could behave incorrectly because PREVRANDAO spans 0 to 2^256, far beyond the 0 to 2^64 range of the old value. That is the broken failure mode: the contract runs but produces wrong results. Separately, PREVRANDAO can be partially manipulated by the validator proposing a given block, because that validator can choose not to reveal a value. This makes it unsuitable as a sole randomness source for high-value applications like lotteries or gaming contracts. A contract relying solely on PREVRANDAO may still function but remains manipulable, which is the unsafe condition.
The broader network architecture also shifted to a dual-client model. Execution clients such as Geth and Erigon continue handling smart contracts and state, while consensus clients such as Prysm and Teku manage network synchronization. The two communicate through the Engine API, an authenticated interface that uses JWT secrets to secure the connection between layers.
A further change with significant implications for application developers was the introduction of new finalized and safe JSON-RPC block tags. Under proof of work, developers queried finality using informal confirmation-count heuristics to estimate whether a transaction was sufficiently settled. Post-Merge, the finalized tag marks a block that has achieved cryptoeconomic finality, while safe marks a block that is unlikely to be reorganized. Applications querying transaction status needed to account for these new tags, which provide a cleaner and more precise mechanism than the counting-based approaches developers had previously relied on.
Regional Stakes: Where Continuity Actually Mattered
The zero-disruption outcome was not trivial in markets where crypto serves as a financial lifeline rather than a speculative vehicle. In Sub-Saharan Africa, on-chain volume reached $100.6 billion between July 2021 and June 2022, rising to $117.1 billion the following year, according to Chainalysis data. About 80% of transfers were under $1,000, and roughly 95% fell into the retail category. Nigeria ranked second globally for crypto adoption in 2023. An unnamed regional consultant quoted in the Chainalysis Sub-Saharan Africa report described crypto in the region as "a rescue, a way to feed their family and solve their daily financial needs."
With Nigeria experiencing inflation above 20% in early 2023, and Ghana hitting 29.8% in June 2022, the stablecoins and DeFi protocols running on Ethereum were active tools for preserving purchasing power. Those services kept running through The Merge without interruption.
South Africa's Financial Sector Conduct Authority implemented a crypto licensing regime in 2022, providing regulatory clarity that coincided with the Merge period and represented a significant step toward formal oversight in the region.
In South Asia, the picture is more complicated. The Central and Southern Asia region received $932 billion in crypto value between July 2021 and June 2022, with India alone accounting for $172 billion. India ranked fourth globally for grassroots crypto adoption that year, and Pakistan ranked sixth. India's DeFi activity had a distinctive profile in this period, with approximately 58% of web traffic to DeFi platforms being NFT-related. However, India's introduction of a 30% flat tax on crypto gains in April 2022, followed by a 1% transaction withholding tax in July 2022, suppressed activity significantly. Both India and Pakistan saw their global adoption rankings decline in the period that followed, showing that infrastructure improvements can be offset by policy decisions at the national level.
On the environmental front, Ethereum's energy reduction matters in South Asian policy circles where ESG-based arguments have been used to restrict crypto. Removing the energy footprint of proof-of-work mining, previously comparable to that of Finland according to analysis cited by Bristows, eliminates one argument regulators in the region have drawn upon.
What Comes Next
The Merge was designed as a prerequisite for further scaling work rather than a scaling solution itself. Consensys noted at the time that the upgrade "paves the way for sharding to improve data availability and bandwidth." The consensus foundation established in September 2022 was intended as the basis for that subsequent work.
For developers building in emerging markets, particularly those working on remittance tools, microfinance applications, or inflation-resistant savings products, the stability of the application layer through that transition meant their work survived intact.