The last mile: how China brought reliable power to everyone, and why generation is the easy part

Bajío Ventures · September 2026

Between 1949 and 2015, China brought electricity to some 900 million people. By the late 1990s it covered 97% of its population; in December 2015, with the last communities in Qinghai, it declared universal access. What came next is what matters most to energy investors: instead of calling the job done, the country spent 830 billion yuan in four years on rural grids alone. Not to connect anyone new, but to make the connection useful.

The story is usually told as a triumph of generation: dams, coal plants, solar parks. It is the other way round. China always knew how to build power plants. What took thirty years and four national programmes was the last mile.

Four waves, one objective

Period	Programme	Investment	What it bought
1998–2002	Rural grid reform and reconstruction (“two reforms, one price”)	288.5 billion yuan	A modernised rural grid and, by 2004, a single residential tariff nationwide except Tibet
2013–2015	Action plan for the population without electricity	24.8 billion yuan (~60% central government)	The last 2.73 million people without power
2016–2019	New round of rural grid upgrades	830 billion yuan (43.5 billion from the central budget)	1.6 million farm wells electrified, 78,000 villages upgraded, 33,000 villages given three-phase supply
2023–2025	Rural grid consolidation	Central budget, rural-grid loan repayment funds and local bonds	Reliability targets of 99.85–99.94%

Sources below. The 2016–2019 figure is the one the NDRC presented in June 2019, with the final year as an estimate.

Three things stand out.

Connection was the cheap part. The last 2.73 million people cost 24.8 billion yuan. Upgrading the grid for those already connected cost more than thirty times that.

The goal shifted from “having power” to “power that works”. The 2016–2019 round was measured on quality: 99.8% supply reliability, 97.9% of voltage within standard and at least 2 kVA of transformer capacity per household. The 2023 plan raised the bar: by 2025, 99.94% reliability, 99.9% voltage compliance and 3.5 kVA per household in the east; 99.85%, 99.2% and 2.3 kVA in the centre, west and northeast.

Three-phase power was an explicit goal. Electrifying 1.6 million irrigation wells and bringing three-phase supply to 33,000 poor villages is not lighting: it gives the countryside the capacity to pump, mill, refrigerate and process. It is productive infrastructure.

What China did differently

It chose technology by density, not ideology. In the final push, connecting a remote household to the grid could cost up to 100,000 yuan (about USD 16,000); a stand-alone solar system, between 7,500 and 11,200. Where there were fewer than 20 households per square kilometre, off-grid solar was used. The result was a mix close to half and half between grid extension and off-grid systems.

It unified the price. Before the reform, electricity in remote parts of Tibet cost more than 2 yuan per kWh; afterwards, 0.54. Since 2004, except in Tibet, the rural residential tariff has been the same as the urban one within each provincial grid. Achieving that requires someone to cross-subsidise the expensive grid with the cheap one. China did it with two state-owned grid companies and earmarked central funds.

It financed the grid as what it is: a public good with slow returns. Of the 2016–2019 spend, only about 5% came from the central budget; the rest came from the grid companies, banks and local governments, with the state standing behind the programme. Nobody waited for each rural line to pay for itself.

Why generation is the easy part

In 2025 China installed 315 GW of solar and some 120 GW of wind in a single year. No country has ever added so much generation so fast. And still, that was never the bottleneck.

In 2016, when wind grew faster than transmission lines, between 17% and 20% of available wind energy was wasted nationally, depending on the source, and 43% in Gansu. In 2025, with solar growing at the same pace, solar curtailment reached 10–17% in Qinghai, Xinjiang and Gansu. Global Energy Monitor sums it up: transmission bottlenecks remain the main constraint.

The state’s answer was, once again, the grid. State Grid invested more than 650 billion yuan in 2025, an all-time record, and plans 4 trillion yuan for 2026–2030, 40% more than in the previous five-year period.

The underlying reason is structural:

• Generation is a product. Modules, inverters, turbines and batteries are standardised equipment, bought on a global market and installed in months. The risk is price and contract risk, and banks know how to measure it.
• The grid is territory. Every substation, right of way and feeder is unique. It depends on local permits, a monopoly operator and a regulated tariff. It is built over years and pays back over decades.
• Generation without a grid is energy you cannot sell. A park that cannot evacuate its output is not an asset; it is a liability with panels.

What it means for Mexico

Mexico does not have the problem China solved in 2015: electricity coverage is already close to universal. It has the problem China has been solving ever since.

CFE’s 2025–2030 plan invests about 3.5 pesos in generation for every peso in the grid. And 46 generation lots were declared void because of grid saturation. It is the same symptom as Gansu in 2016: generation ready to go, not enough lines.

For investors in the Bajío, the practical reading is threefold:

1. Don’t compete where everyone competes. Building a solar park is the part the market already knows how to do. The scarcity is in connection capacity.
2. Invest where the load is. Behind-the-meter generation and storage, next to the industrial plant, avoids the saturated grid instead of waiting for it. It is the Mexican version of China’s density rule: use the local solution where the grid will not arrive in time.
3. Measure what China measures. For a plant, what matters is not having a contract with CFE, but reliability, voltage quality and available substation capacity. Those indicators now decide where a production line can be installed.

It took China thirty years to learn that the last mile is the one that costs. Mexico can learn it faster, and the capital that understands it first will be on the right side of the bottleneck.

Sources: He and Victor, Experiences and lessons from China’s success in providing electricity for all, 2017 · Business Standard/PTI, China achieves universal power access, Dec 2015 · Rural Electrification (农村电气化), 40 years of rural power development in China, Nov 2018 · Government of China, State Council policy briefing on rural grid upgrades, 28 Jun 2019 · Government of China, China steps up upgrade of rural power grids, Jul 2019 · NEA, rural grid upgrades do not raise rural tariffs, Jul 2019 · NDRC, NEA and National Rural Revitalization Administration, guidance on rural grid consolidation, Jul 2023 · pv magazine, China adds 315 GW of solar in 2025, Jan 2026 · Global Energy Monitor, China Wind & Solar brief, Jul 2026 · Physics World, green power curtailment in China · Yicai Global, State Grid to invest record 4 trillion yuan over five years

Bajío Ventures — a Mexican energy-infrastructure investment group based in Guanajuato: capital, a data platform on the region’s industrial base, and a German industrial channel, with its own advisory presence in Berlin.